WarpOS released !!!
Steffen Haeuser
Hi!
Well, some of you probably already have heard that the ppc.library of Phase 5
of the PowerUP Boards causes problems, when you switch often between 68k and
PPC CPU, causes great slowdowns. Especially for games and emulators this is
quite problematic. But it is a software problem, not a hardware problem.
The Firm Haage&Partner, known for their StormC compiler, have now done
something about it. They developped an OS expansion called "WarpOS - The
WarpSpeed Solution", which replaces the functionality of ppc.library and
partially also that of exec.library. WarpOS is mainly coded in PPC Assembler,
and it is VERY fast.
Some of the features WarpOS provides:
- PPC Native Multitasking functions
- Memory Protection
- Fast 68k/PPC Communication and Context-Switching
- PPC Native Signalling and Messaging
- Special Support for MMU and Caches (You can manipulate them with some
nice library functions)
- Tutorial about PPC Game Coding
- Demos: Voxelgraphics demo and fractal program and some more stuff
- Extra stuff: The first PPC c2p algorithm and some PPC Memory-Copy Functions
- Benchmark program
- Most functions will be automatically handled by StormC
- ...
The only disadvantage of WarpOS is, that when you start a WarpOS program,
ppc.library using programs won't run any longer (until the next reset). But
well, i expect most developpers will use WarpOS anyways. Especially because it
is much faster. Haage&Partner said, especially about game coding "game coding
under PowerUP simply needs WarpOS, the ppc.library just causes too many delays
when doing a context-switch".
Well, WarpOS seems to be an easy to use and efficient solution. It is Freeware
and will probably soon show up on the Unofficial PowerUP Site at
http://hem.passagen.se/studiox/powerup/index.html. It also is already found at
my BBS telnet 194.123.83.1, login WarpUP, Password WarpUP.
I expect it to showup on Aminet and on http://www.haage-partner.com soon, too,
of course...
In the docs there is written that the Voxelgraphics-Demo runs on the PPC603e
about 3 times faster than on a 50 MHz 060 CPU !!! And on the PPC604e it will
run even faster.
Steffen Haeuser
Mathias Ortmann
Wolf Dietrich <w...@gf.phase5.de> wrote:
[...]
>Furthermore, phase 5 digital products will not tolerate that an
>incompatible software takes over the control of the PowerUp hardware and
>prevents the user from running software that has been developed for the
>PowerUp boards by phase 5 digital products as well as many other
>vendors.
[...]
This is ironic. Even Microsoft pales beside this kind of business practice.
The following quote from Ralph Schmidt completes the picture:
<Laire> sotto: warpup mag vielleicht auf den bis jetzt ausgelieferten boards
laufen...das wird sich ab uebernaechste woche aendern
(Translation: "WarpUP may be compatible with the current boards, but this
will change in a week or two.")
--
Mathias Ortmann ort...@informatik.tu-muenchen.de
UAE 0.6.9 Win32/DirectX @ http://www.informatik.tu-muenchen.de/~ortmann/uae/
Wolf Dietrich
Mathias Ortmann wrote:
>
> Wolf Dietrich <w...@gf.phase5.de> wrote:
>
> [...]
>
> >Furthermore, phase 5 digital products will not tolerate that an
> >incompatible software takes over the control of the PowerUp hardware and
> >prevents the user from running software that has been developed for the
> >PowerUp boards by phase 5 digital products as well as many other
> >vendors.
>
> [...]
>
> This is ironic. Even Microsoft pales beside this kind of business practice.
No. As it is known from your contributions to german-language
newsgroups, you are mixing up things and argueing against phase 5 with
limited knowledge and understanding of the facts.
phase 5 supplies, as I stated, a solution which consists of hardware and
software. We provide a HAL on which every developer can integrate any
solution he likes, while staying compatible. As far as the PowerUp
boards are concerned, we are the system vendor, and believe that we
should control the system software that controls the hardware underneath
the HAL, especially to make sure that all future developments by all the
developers who support PowerUp remain compatible.
If you want to make a comparison with Microsoft, you should question the
strategy of Haage&Partner, who - as a software and development system
vendor - want to take over full control of the hardware that we deliver.
That means that the kernel, the (only compatible) compiler as well as
some application software - including their java development - come from
one source. They even claim their software could be a standard for
future PPC Amigas from Amiga Inc. - provided that Amiga Inc. let
themselves and their OS development be controlled by the H&P solution,
what H&P doesn't say.
Now why has Haage&Partner spent so many efforts in developing an
incompatible kernel, which is quite similar to our own system software
and offers no relevant advantages? Could it be that they want to control
the way any future development will go (and sell the software that is
mandatory for the development)? Or do you think that instead of
developing applications which they could sell, they spend their efforts
developing a kernel and give it away for free, just because of idealism?
Everybody may draw an own conclusion or do the comparison as in the
paragraph before...
Also, your translation was wrong. Ralph Schmidt didn't say "WarpUP may
be compatible with the current boards, but this will change in a week or
two" but said "WarpUp may run on the boards shipped so far, but this
will change in two weeks from now." The difference is small, but
important: WarpUp HAS NEVER BEEN COMPATIBLE, but runs only because
Haage&Partner can currently disable the PowerUp system software. As in
future revision of PowerUp the PPC-library and other parts of our
software will be located in the Flash-ROM (a feature that had been
planned from the beginning regardless of the current haage&partner
situation), WarpUp will not longer run as it is incompatible to our
solution.
Please be more accurate when you are translating statements in the
future.
------------------------------------------------------------
Wolf Dietrich, General Manager phase 5 digital products
Return mail address: w...@gf.phase5.de http://www.phase5.de
------------------------------------------------------------
Mathias Ortmann
Wolf Dietrich <w...@gf.phase5.de> wrote:
>Mathias Ortmann wrote:
>>
>> Wolf Dietrich <w...@gf.phase5.de> wrote:
>>
>> [...]
>>
>> >Furthermore, phase 5 digital products will not tolerate that an
>> >incompatible software takes over the control of the PowerUp hardware and
>> >prevents the user from running software that has been developed for the
>> >PowerUp boards by phase 5 digital products as well as many other
>> >vendors.
>>
>> [...]
>>
>> This is ironic. Even Microsoft pales beside this kind of business practice.
[...]
>phase 5 supplies, as I stated, a solution which consists of hardware and
>software. We provide a HAL on which every developer can integrate any
>solution he likes, while staying compatible. As far as the PowerUp
>boards are concerned, we are the system vendor, and believe that we
>should control the system software that controls the hardware underneath
>the HAL, especially to make sure that all future developments by all the
>developers who support PowerUp remain compatible.
This answers the questions regarding AmigaOS 4.0 or pOS on phase5-controlled
PPC boards that people post here once in a while...
Paul Kolenbrander
On Sat, 27 Sep 97, Steffen Haeuser typed the following line(s):
[...]
> something about it. They developped an OS expansion called "WarpOS
^^^^
[...]
Uhoh! They'll probably be visited by men in suits shortly. As IBM
uses WARP for it's OS/2 3.0 and as far as I know, has it TM'd. It
may be advisable to check the legality of 'warpOS' now instead of
later...
Take care,
--__
/_/ |/ E-mail: paul-at-serena.iaehv.nl
/aul |\olenbrander WWW : http://www.iaehv.nl/users/paul/
-o-o-o-o-o-o-o-o-o-o-o-o-o-o-o-o-o-o-o-o-o-o-o-o-o-o-o-o-o-o-
- NOTE: Replace -at- in email addy with @. Do NOT use Reply -
US$500 handling charge applied to unsolicited commercial mail
Wolf Dietrich
This is a short preliminary statement, which will be extended during the
next week and can then be found on our web site http://www.phase5.de.
phase 5 digital products is shipping the PowerUp boards as a complete
solution, consisting of the powerful dual-processor hardware and the
comprehensive PPC library and other stuff. The PPC library is a complete
PPC native multitasking and multiprocessing kernel. It features
everything that is necessary for all kind of software developments, even
for adoption of different or future OS versions. The information about
disadvantages and slow communication between the two CPUs is nonsense,
as our system software provides a comprehensive message and signalling
system, which even provides protectable memory areas for each task,
making PowerUp-compliant software ready for future memory-protected OS
updates or versions.
AS A MATTER OF FACT, THE FUNCTIONALITY WHICH HAAGE&PARTNER OUTLINES AND
PROMOTES IS NOTHING SPECIAL, BUT A STANDARD FUNCTIONALITY PROVIDED BY
THE PHASE 5 POWERUP SYSTEM SOFTWARE. We simply did not make so much
noise about it. It is also nonsense that the WarpUp software by
Haage&Partner is faster then the PowerUp system software from phase 5
digital products; this may only be true when StormC is used for the
compilation a PowerUp application.
The WarpUp from Haage&Partner is a solution which is not compatible to
the phase 5 solution. As it does not use the HAL that is provided by
phase 5 digital products, but hacks the hardware by using an incomplete
knowledge of the hardware functionality - which has been aquired by
reengineering the phase 5 digital products software - it is most likely
that it will not work with future revision changes or different product
releases of the PowerUp series.
Furthermore, phase 5 digital products will not tolerate that an
incompatible software takes over the control of the PowerUp hardware and
prevents the user from running software that has been developed for the
PowerUp boards by phase 5 digital products as well as many other
vendors. All software which is supplied by phase 5, and in which we have
invested lots of work and financial efforts to make the PowerUp a useful
product, would be disabled once the Haage&Partner software is installed
in the system. That includes for example the fast CyberGL native library
as well as the MPEG video and audio decoding integrated into CGX V3.
Consequently, third party products using such advanced functionality
would not run. USAGE OF THE THE HAAGE&PARTNER SOLUTION WOULD RESULT IN A
COMPLETE INCOMPATIBILITY MESS. In the interest of our customers, who buy
our products with high expectations, and all our developers who expect a
completely compatible product line, we will take care that the phase 5
system software can not be disabled.
Haage&Partner has - from the beginning - rejected to use the ELF format
which phase 5 has introduced on the PowerUp boards. The reason to choose
this widely accepted format was to open the Amiga/PowerUp development to
industry software standards. As an example, the ELF format allows
professional development for PowerUp even on other platforms (such as
Motorola PPC systems, IBM AIX systems, even on Sun or under WindowsNT),
where high performance PPC Cross compilers are available, which
integrate far superior experience and performance; phase 5 is currently
testing such compilers which may offer the chance to accelerate software
such as the CyberGL native library and the PPC MPEG library even
further. Additonally, soon an update for SAS/C on the Amiga can be
expected, which supports PowerUp and the ELF format.
Haage&Partner, instead, wants to establish their own "extended
Hunk-format", which may is derived from the Amiga Hunk format, but still
a new and proprietary solution. It is a concept that forces developers
to use the only solution supporting this format - the Storm C compiler
by Haage&Partner. Haage&Partner declares it's product being open and
Amiga-compliant - in fact, it is only open to the Haage&Partner standard
and compliant to their own software solutions.
At this point, it is important to mention that the software concepts of
phase 5 digital products have all the time been focused on developments
that open doors and smooth upgrade paths to future and completely
revised OS versions, which incorporate functionalities such as memory
protection and multi-processing (which is not supported by the
Haage&Partner concept, BTW). This has e.g. also been proposed to Amiga
International/Gateway2000 in May this year, together with our offer for
extensive technological cooperation. To reach such a goal, we feel it is
mandatory to set today guidelines for programming that make current
Amiga/PowerUp developments more independent from the current OS revision
by clear structured and object-oriented programming, instead of using
the StormC concept of simply recompiling older sources (this issue will
be adressed in detail later).
As the solution from Haage&Partner is not compatible and can - because
of it's proprietary software concept - not fullfill demands of
professional developers, it is very unlikely that developers will choose
it. phase 5 digital products highly recommends not to use this
incompatible system, as well as the Storm C compiler which - because of
the concept - is wasting most of the performance that the PowerUp boards
can deliver (this issue will also be adressed in detail later).
phase 5 digital products has supported Haage&partner with free developer
systems, as well as all necessary information to develop compatible
software and tools, thus enabling Haage&Partner to become a leading
software vendor. We also held many meetings explaining our strategy and
goals to make the PowerUp a useful platform not only for a new
generation of high-performance applications, but also for the
development towards a new OS or OS version that fullfills the demands of
the next century. To our surprise and disappointment Haage&Partner has
mainly used their status to create a replacement for the PowerUp system
software which is intended to set a standard controlled by
Haage&Partner, and they even attempt to trash the complete and
comprehensive software development in which we have spend many efforts.
At the same time, they have yet not been able to present a stable
running PowerPC application which the Amiga/PowerUp user would have an
advantage from. We highly disregard this contraproductive and
confrontational behaviour, and will completely drop any support for
Haage&Partner as a PowerUp developer as well as for any of their
products.
Steffen Haeuser
˙° Mag...@BIRDLAND.es.bawue.de schrieb : ˙´
Hi!
An update as to WarpOS: As WarpOS was for a long time handled a "secret" by
Haage&Partner, and then suddenly released, i assumed, that the "secret" was
because H&P had to clear up things with Phase 5. Especially i assumed, they
now (as it was no longer a secret) cleared up things with Phase 5.
It seems the problem is not solved. Phase 5 (for what reasons i don't know)
still do not like WarpOS. After my posting i was contacted by the coder of the
original ppc.library (who did not like this posting also). I now asked him in
the reply for some answers about his view of things... i hope i can clear up
things, soon... i will talk tomorrow again with Mr. Haage, too.
But, given the speed comparision of both software systems, i would not know
what speaks against WarpOS. It is hardware-independent, BTW, so it will always
run on future Amiga PPC Hardware, if this is now a PPC Board or something new
from AT (like outlined on the Haage&Partner WWW Site).
Of course, as the coder of ppc.library outlined, it is not completely the
fault of the software... 68k/PPC switch is a hardware problem, also. But
appearently, it can be done much faster than ppc.library does it.
Well, i hope i will get as much info from Phase 5 like from Haage&Partner. I
would like to remain a neutral view.
BTW, the coder of the ppc.library told some interesting news: There
appearently will be a PPC Version of SAS/C as free upgrade !!!
Steffen Haeuser
James McArthur
[ .. ]
: This is ironic. Even Microsoft pales beside this kind of business practice.
M$ dont make hardware, only software..
: The following quote from Ralph Schmidt completes the picture:
: <Laire> sotto: warpup mag vielleicht auf den bis jetzt ausgelieferten boards
: laufen...das wird sich ab uebernaechste woche aendern
: (Translation: "WarpUP may be compatible with the current boards, but this
: will change in a week or two.")
They make the hardware, they should be able to say what runs and what doesnt.
Seems kind of silly thought that H&P would stab P5 in the back since P5 are
currently the only Amiga PowerPC vendors.. makes me wonder what we _arent_
being told.. (cue conspiracy theory.. :>)
--
Amiga 4000/060/PPC150 CV64/3D 14Mb James McArthur
Currently reposting ALL spam to:
*@iemmc.org *@answerme.com *@global-impact.com *@cafes.net
*@mail.enterprise.net
Rudi Chiarito
Wolf Dietrich (w...@gf.phase5.de) wrote:
>boards are concerned, we are the system vendor, and believe that we
>should control the system software that controls the hardware underneath
>the HAL, especially to make sure that all future developments by all the
>developers who support PowerUp remain compatible.
Commodore was the system vendor of Amiga. They didn't care if someone else
wrote other system software that controled the hardware, like Minix, BSD, etc.
If a customer buys an hardware product, I guess he/she has the RIGHT to
choose what to run on it.
>Now why has Haage&Partner spent so many efforts in developing an
>incompatible kernel, which is quite similar to our own system software
>and offers no relevant advantages? Could it be that they want to control
There ARE advantages. I saw a demo (and I had an interview with H&P for a
magazine) last week running on both libraries and the performance boost sported
by WarpOS was noticeable. And that was on a mere 603.
--
"Confusion will be my epitaph As I crawl a cracked and broken path"
Rudi Chiarito - ATO, Team Amiga and EpOS - Jay Miner Society founding member
EMail: chia...@cli.di.unipi.it - WWW: http://www.cli.di.unipi.it/~chiarito/
MISTAKES/MISSPELLINGS ARE FICTIONAL: A SIMILARITY TO REAL ONES IS INCIDENTAL
Wolf Dietrich
Rudi Chiarito wrote:
>
> >Now why has Haage&Partner spent so many efforts in developing an
> >incompatible kernel, which is quite similar to our own system software
> >and offers no relevant advantages? Could it be that they want to control
>
> There ARE advantages. I saw a demo (and I had an interview with H&P for a
> magazine) last week running on both libraries and the performance boost sported
> by WarpOS was noticeable. And that was on a mere 603.
Could it be that the demo showing the "advantages" has been provided by
Haage&Partner? Could it be that you have seen it running on a Beta
developer board (the answer is "yes" as we didn't ship any 603 release
products) with probably an older beta version of our PowerUp system
software running?
I assume that you have seen "advantages" of a Storm C compiled program
running under WarpUp compared to the Storm C compiled version which is
compatible with the PowerUp system software. This is a useless
comparison, because software which is developed for PowerUp with Storm C
is wasting most of the possible performance of the PowerUp boards due to
it's "quick-and-dirty-conversion" concept.
Rudi Chiarito
Wolf Dietrich (w...@gf.phase5.de) wrote:
>Could it be that the demo showing the "advantages" has been provided by
>Haage&Partner?
Of course, you can't expect a demo of WarpUp from anyone else but them ;)
>I assume that you have seen "advantages" of a Storm C compiled program
>running under WarpUp compared to the Storm C compiled version which is
>compatible with the PowerUp system software. This is a useless
>comparison, because software which is developed for PowerUp with Storm C
>is wasting most of the possible performance of the PowerUp boards due to
>it's "quick-and-dirty-conversion" concept.
Ok.. what is "most of the possible performance" lost for, then? What are
they doing so terribly wrong to get such a performance loss? I'm trying to
get some detailed FACTS, but all I see are vague statements.
Wolf Dietrich
Rudi Chiarito wrote:
>
> >Could it be that the demo showing the "advantages" has been provided by
> >Haage&Partner?
>
> Of course, you can't expect a demo of WarpUp from anyone else but them ;)
Yes, and the demo program running under the PowerUp System Software has
also been provided by them. Oh, by the way, was it the "Set Pixel"-Demo
which they have shown you? If so, this demo is as useful as testing a
Porsche by driving backwards.
> >I assume that you have seen "advantages" of a Storm C compiled program
> >running under WarpUp compared to the Storm C compiled version which is
> >compatible with the PowerUp system software. This is a useless
> >comparison, because software which is developed for PowerUp with Storm C
> >is wasting most of the possible performance of the PowerUp boards due to
> >it's "quick-and-dirty-conversion" concept.
>
> Ok.. what is "most of the possible performance" lost for, then? What are
> they doing so terribly wrong to get such a performance loss? I'm trying to
> get some detailed FACTS, but all I see are vague statements.
>
I am not a software developer, and can just explain it with the basic
understanding that I have, so please excuse if this is not going too
much into the details. We at phase 5 digital products do recommend that
software developers who want to support PowerUp rewrite their programs
in a way that they build software modules in form of tasks which run on
PPC and 68k. In this way, the software is well structured, open for
future developments and object oriented. A simple program for example
could consist of a 68k task which handles the GUI and OS calls, and
calls PPC tasks for executing operations which shall be accelerated by
using the PPC. As we provide comprehensive functionality and a very fast
new message system in our PowerUp System Software, this solution is very
efficient, and it allows full parallel operation of multiple CPUs; in
the example above, the 68k task could continue with execution of
commands while one or more PPC tasks are busy with other stuff, and
wouldn't need to wait until the PPC task(s) returns (unless, of course,
the 68k task would need the result of the PPC task(s) to continue it's
work).
By using such a structured and object oriented programming approach, the
software is also much more easily updatable. If e.g. GUI and OS calls
would change in future revisions of an OS, the developer would only need
to replace the module that handles the GUI and the OS calls, and
continue to use the other modules.
H&P have been promoting their Storm C with a concept that you simply
recompile the existing C code which has been written for 68k Amiga and
tell the compiler to generate PPC code. This may be very comfortable for
the developer, but is completely unefficient. Not only that in the worst
case you have a random mix of 68k and PPC code, but also the two CPUs
will not run in parallel as it is one continuous program flow which
jumps between PPC and 68k. If a linear code sequence is executed, the
68k and the PPC always work one after the other, in the worst case with
lots of switches and a useless overhead.
A significant loss of performance is the result of this way of porting
applications to the PowerPC. It could result in a scenario where
developers use Storm C, recompile their code, find that it is two times
faster than on a 68060 and decide that this is OK, while with some
thinking and conceptual consideration the program may would work five or
ten times faster than on the 68060. Insofar this is a poor solution from
which maybe a less motivated developer would benefit, being enabled to
do a quick-and-dirty conversion, but not the customer, who would get
products that do not exploit the performance of his hardware at all.
Paul Cardwell
Paul Kolenbrander wrote:
> Uhoh! They'll probably be visited by men in suits shortly. As IBM
> uses WARP for it's OS/2 3.0 and as far as I know, has it TM'd. It
> may be advisable to check the legality of 'warpOS' now instead of
> later...
...and Paramount .
Paul
Espen Berntsen
On Sun, 28 Sep 1997 16:04:15 +0200, Wolf Dietrich <w...@gf.phase5.de>
wrote:
>Furthermore, phase 5 digital products will not tolerate that an
>incompatible software takes over the control of the PowerUp hardware and
>prevents the user from running software that has been developed for the
>PowerUp boards by phase 5 digital products as well as many other
Say what? A new bogged down hardware key to prevent other from making
things for the new hardware? Like M$ did for NT when they used
undocumented lib calls to make email programs (and thus preventing
others "from taking over the mail programs developed by M$ as well as
many other vendors".
>vendors. All software which is supplied by phase 5, and in which we have
>invested lots of work and financial efforts to make the PowerUp a useful
>product, would be disabled once the Haage&Partner software is installed
>in the system. That includes for example the fast CyberGL native library
>as well as the MPEG video and audio decoding integrated into CGX V3.
So if it finds some H&P stuff you disable what you can? How nice.
Remind me of not getting the alleged A\Box.
>Haage&Partner, instead, wants to establish their own "extended
>Hunk-format", which may is derived from the Amiga Hunk format, but still
>a new and proprietary solution. It is a concept that forces developers
>to use the only solution supporting this format - the Storm C compiler
>by Haage&Partner. Haage&Partner declares it's product being open and
>Amiga-compliant - in fact, it is only open to the Haage&Partner standard
>and compliant to their own software solutions.
And the PowerUp board is only compliant with P5 stuff. And the big
difference here is???
--
Nameless <> Name...@mo.himolde.no
Let's face it, if there is one thing that's addictive: it's politics. And
politics, unlike pot, most definitely _does_ cause brain damage.
Espen Berntsen
On 29 Sep 1997 07:48:43 GMT, Spamm...@it.ntu.edu.au (James
McArthur) wrote:
>: The following quote from Ralph Schmidt completes the picture:
>
>: <Laire> sotto: warpup mag vielleicht auf den bis jetzt ausgelieferten boards
>: laufen...das wird sich ab uebernaechste woche aendern
>
>: (Translation: "WarpUP may be compatible with the current boards, but this
>: will change in a week or two.")
>
>They make the hardware, they should be able to say what runs and what doesnt.
Are you nuts? What if they go bankrupt? What if I want to run some
other OS? What if I install some OTHER hardware? Seems to me that they
have taken all teh bad sides of the Amiga and magnified it.
Mans Engman
On Mon, 29 Sep 1997 19:52:31 GMT, Espen Berntsen <name...@yme.himolde.no>
wrote:
>>incompatible software takes over the control of the PowerUp hardware and
>>prevents the user from running software that has been developed for the
>>PowerUp boards by phase 5 digital products as well as many other
>Say what? A new bogged down hardware key to prevent other from making
>things for the new hardware? Like M$ did for NT when they used
>undocumented lib calls to make email programs (and thus preventing
>others "from taking over the mail programs developed by M$ as well as
>many other vendors".
Ehh, I don't see an analogy. An analogy would be that someone develops
a Windoze application, that when run shuts down all other Windoze
programs, and M$ refuses to give this program a "M$-Windoze-approved"
sticker.
>>vendors. All software which is supplied by phase 5, and in which we have
>>invested lots of work and financial efforts to make the PowerUp a useful
>>product, would be disabled once the Haage&Partner software is installed
>>in the system. That includes for example the fast CyberGL native library
>>as well as the MPEG video and audio decoding integrated into CGX V3.
>
>So if it finds some H&P stuff you disable what you can? How nice.
>Remind me of not getting the alleged A\Box.
No, no, no. Running WarpOS means that P5 software, or software created
with P5's _free_ devkit won't run anymore. Such a solution is a big step
backwards, back to the good old days when you had to reset the computer
between running programs.
On the other hand I should note that WarpUP contains two versions
of powerpc.library -- one that seems to be compatible with ppc.library
(a layer "on top" of it), and one version using WarpOS which will
grab the ppc board completely. So programs can be made to work for
both cases. But any program that needs WarpOS-specific features needs
the right library to run, and will thus disable the possibility to run
ELF binaries at the same time. IMHO, stupid.
Some assorted thoughts:
The H&P solution is to run "mirror tasks" -- for each task on the m68k
there is a corresponding PPC task, and the program can transfer its
control flow between those, so usually only one of them runs at the
same time. As H&P have pointed out, this can make the task of porting
an applications quite easy, since the control-transfer can be
generated automagically (with storm-c ;). On the other hand, this
is not a generally efficent approach because of the "context-switches".
I guess that's the reason why they hack a kernel which makes this work
better.
>>Haage&Partner, instead, wants to establish their own "extended
>>Hunk-format", which may is derived from the Amiga Hunk format, but still
>>a new and proprietary solution. It is a concept that forces developers
>>to use the only solution supporting this format - the Storm C compiler
>>by Haage&Partner. Haage&Partner declares it's product being open and
>>Amiga-compliant - in fact, it is only open to the Haage&Partner standard
>>and compliant to their own software solutions.
>And the PowerUp board is only compliant with P5 stuff. And the big
>difference here is???
Developing the P5 way means; both free compilers and tools (gcc etc),
and also (possibly) the choice to use other industry-standard tools that
can work with ELF objects. Now tell me, what way other than the commercial
Storm-tools is there to compile his/her applications to the H&P format?
The WarpUp docs don't give a clue, anyhow.
/Mans
mo...@lysator.liu.se
Rene Laederach
Hello Wolf!
Wolf Dietrich typed this on Sun, 28 Sep 1997 17:48:21 +0200 about 'Re:
WarpOS released !!!':
WD> that we should control the system software that controls the hardware
WD> underneath the HAL, especially to make sure that all future
WD> developments by all the developers who support PowerUp remain
WD> compatible.
Hmmm....I think this is also a way which is more or less rotten.
Have a comitee. Have lots of firms working with you. Get as many developers as
possible (DKB? NewTek?) work with you. This way, you can enforce standards much
better. Yet nobody will be left out in the cold. Just imagine software engineers
porting PowerAnimator to a competing PPC Amiga system, but not yours. You will
loose sales, because PowerAnimator seems to be a highly cool software. As well
as Maya, and some other Alias/Wavefront stuff. If there's only one underlying
system, but multiple vendors, you might be the lucky Amiga firm getting the
sales of nice multicpu machines required for that work.
Amiga users are highly allergic against monopolies and would evade either
solution which is being monopolized by any Amiga firm, be it you or
Haage&Partner.
I'm slowly tired of nitty-picky infighting in the Amiga software&hardware
industry as it is of now. Just see the P96<->CyberGFX fight with the whole play
on the V43 datatype. If you want an Amiga market, you should rather follow the
motto "Do not ask what the Amiga can do for you, but what you can do for the
Amiga.".
You can do a lot if you just wanted, do you?
--
FIDO: 2:301/133 & 135 | Member We're returning!
Internet mu...@snoop.alphanet.ch | Team AMIGA - the true avantgarde
Holger Metz
Mans Engman schrieb:
> Ehh, I don't see an analogy. An analogy would be that someone develops
>
> a Windoze application, that when run shuts down all other Windoze
> programs, and M$ refuses to give this program a "M$-Windoze-approved"
> sticker.
then there will be a phase5 sticker on every software :) phase5 approved
! :)
> No, no, no. Running WarpOS means that P5 software, or software created
>
> with P5's _free_ devkit won't run anymore. Such a solution is a big
> step
> backwards, back to the good old days when you had to reset the
> computer
> between running programs.
>
hmm, the best way would be for phase5 and H&P that they come together
for one solution ! Awar would be the last this small amiga community
would help !
I think the best way would be the user/developer jugdes what is best for
you and installs "his"
software he wants !!
> On the other hand I should note that WarpUP contains two versions
> of powerpc.library -- one that seems to be compatible with ppc.library
>
> (a layer "on top" of it), and one version using WarpOS which will
> grab the ppc board completely. So programs can be made to work for
> both cases. But any program that needs WarpOS-specific features needs
> the right library to run, and will thus disable the possibility to run
>
> ELF binaries at the same time. IMHO, stupid.
> Some assorted thoughts:
> The H&P solution is to run "mirror tasks" -- for each task on the m68k
>
> there is a corresponding PPC task, and the program can transfer its
> control flow between those, so usually only one of them runs at the
> same time. As H&P have pointed out, this can make the task of porting
> an applications quite easy, since the control-transfer can be
> generated automagically (with storm-c ;). On the other hand, this
> is not a generally efficent approach because of the
> "context-switches".
> I guess that's the reason why they hack a kernel which makes this work
>
> better.
>
they didnt hack a kernel.. they made a message system like phase5
library ! ..Thats the reason for the SPEED increase.. their own
programms in elf run slower because
they use a ppc.library from december 96.. which has this task
switching..
this is the reason why their prog. runs so slow ! with the newer
ppc.library it would run
possibly the same speed !
>
>
> Developing the P5 way means; both free compilers and tools (gcc etc),
> and also (possibly) the choice to use other industry-standard tools
> that
> can work with ELF objects. Now tell me, what way other than the
> commercial
> Storm-tools is there to compile his/her applications to the H&P
> format?
> The WarpUp docs don't give a clue, anyhow.
>
they give a clue !! you have includes for all assembler and c compiler
!! THE MAIN problem isthat there are only 2 compilers which produce PPC
.. Storm C and GCC .. ! Does this give you a clue..?
I think the main reason why phase5 would like all developers stick to
their card is, that if they dont, they would loose the PPC Market..
Warpos doesnt stick to the Hardware.. (they do something similar as the
ISDN Cards on PeCe. One fossil driver for the
card itself on the Hardwareside and one software (C)API for the
developers..) so when everyone uses this WarpOS System, other companys
could produce PowerPC cards which runs with every software produced for
Warpos!! But if phase5 does something against it.. no other company
could produce powerpc cards, because they had to be compatible with the
P5 cards.. (Ralph Schmidt library runs only on phase5 cards) or the
software you bought (which uses p5 ppc.library) wont run... its the same
as microsoft does with their visual products.. you can only produce
software which runs under Microsoft OS. But looking on the other side of
H&P its nearly the same... when some company wants to produce a powerpc
card they have to licence warpup from H&P (or write a driver).
So to come to a end... the best way it would be to let the user decide
what software he uses.. or what would YOU
do if you buy a pc and get told that you should ONLY INSTALL windows95
on it or the Hardware wont work ??
> /Mans
> mo...@lysator.liu.se
Stefan Burstrom
On 29-Sep-97 20:52:31 Espen Berntsen (name...@yme.himolde.no) wrote:
>Say what? A new bogged down hardware key to prevent other from making
>things for the new hardware? Like M$ did for NT when they used
>undocumented lib calls to make email programs (and thus preventing
>others "from taking over the mail programs developed by M$ as well as
>many other vendors".
As long as software stay compatible anyone can do what they want. H&P's
solution is neither compatible nor very efficient.
>>vendors. All software which is supplied by phase 5, and in which we have
>>invested lots of work and financial efforts to make the PowerUp a useful
>>product, would be disabled once the Haage&Partner software is installed
>>in the system. That includes for example the fast CyberGL native library
>>as well as the MPEG video and audio decoding integrated into CGX V3.
>So if it finds some H&P stuff you disable what you can? How nice.
>Remind me of not getting the alleged A\Box.
Why do you look at P5 as the devil? I don't find H&P any better at all, on
the contrary...!
>>Haage&Partner, instead, wants to establish their own "extended
>>Hunk-format", which may is derived from the Amiga Hunk format, but still
>>a new and proprietary solution. It is a concept that forces developers
>>to use the only solution supporting this format - the Storm C compiler
>>by Haage&Partner. Haage&Partner declares it's product being open and
>>Amiga-compliant - in fact, it is only open to the Haage&Partner standard
>>and compliant to their own software solutions.
>And the PowerUp board is only compliant with P5 stuff. And the big
>difference here is???
The difference is that the PowerUp software is based on industri standard,
has remained compatible with the initial releases given out to developers.
H&P's solution disables the PowerUp software and makes 100% compliant
PowerUp software break.
regards, Stefan Burstroem
-----------------------------------------------------------------------
>> Irl: Stefan Burstroem << >> Omnipresence Intl. << >> Irc: Yabba <<
>> Phone: +46 (0)46-211 40 84 << >> EMail: ste...@omnipresence.com <<
-----------------------------------------------------------------------
ps. The expressed opionins above are not P5's nor Omnipresence Intl's.
It is just my expressed opinion based on months of discussions with
Phase5 about the PowerUp system and by looking at the StormC solution.
George Noel
In article <342E63DF...@gf.phase5.de>,
[SNIP]
>The WarpUp from Haage&Partner is a solution which is not compatible to
>the phase 5 solution.
Well that is obvious, as was also stated.
> As it does not use the HAL that is provided by
>phase 5 digital products, but hacks the hardware by using an incomplete
>knowledge of the hardware functionality - which has been aquired by
>reengineering the phase 5 digital products software - it is most likely
>that it will not work with future revision changes or different product
>releases of the PowerUp series.
This was what I was worried about in the first place when it came to
PowerUP but it was pointed out, that PhaseV made sure things will work on
the various CPUs and not be specific to only certain ones. So yes, if this
is the case, it is a major reason to stay away from WarpOS.
>Furthermore, phase 5 digital products will not tolerate that an
>incompatible software takes over the control of the PowerUp hardware and
>prevents the user from running software that has been developed for the
>PowerUp boards by phase 5 digital products as well as many other
>vendors.
Now THIS is scary. I do not like that attitude.
> All software which is supplied by phase 5, and in which we have
>invested lots of work and financial efforts to make the PowerUp a useful
>product, would be disabled once the Haage&Partner software is installed
>in the system.
Sure.. so what? A reboot could send it back if it is done properly.
[SNIP]
>would not run. USAGE OF THE THE HAAGE&PARTNER SOLUTION WOULD RESULT IN A
>COMPLETE INCOMPATIBILITY MESS.
But could provide some competition, no?
> In the interest of our customers, who buy
>our products with high expectations, and all our developers who expect a
>completely compatible product line, we will take care that the phase 5
>system software can not be disabled.
Again, that is scary. And not wanted.
[SNIP]
>further. Additonally, soon an update for SAS/C on the Amiga can be
>expected, which supports PowerUp and the ELF format.
Hey.. this is great news!
>Haage&Partner, instead, wants to establish their own "extended
>Hunk-format", which may is derived from the Amiga Hunk format, but still
>a new and proprietary solution. It is a concept that forces developers
>to use the only solution supporting this format - the Storm C compiler
>by Haage&Partner.
But instead, you also want developers to ONLY use your solution so what
makes you any better in that regard?
> Haage&Partner declares it's product being open and
>Amiga-compliant - in fact, it is only open to the Haage&Partner standard
>and compliant to their own software solutions.
Geez.. so is yours. Point moot.
>As the solution from Haage&Partner is not compatible and can - because
Well so far it is your word against theirs.
>of it's proprietary software concept - not fullfill demands of
>professional developers,
I say let them speak!
> it is very unlikely that developers will choose
>it.
Again, let them speak for themselves.
> phase 5 digital products highly recommends not to use this
>incompatible system,
That is your choice but you cannot enforce it.
> as well as the Storm C compiler which - because of
>the concept - is wasting most of the performance that the PowerUp boards
>can deliver (this issue will also be adressed in detail later).
I'd be interesting in reading that but for now, that recommendation was
not needed and actually by the sounds of some of these talks, unless these
can be proven, it sounds liable for damages to me.
But as my personal opinion about what I know about StormC and SAS/C, I'd
got with SAS/C anyday.
>phase 5 digital products has supported Haage&partner with free developer
>systems, as well as all necessary information to develop compatible
>software and tools, thus enabling Haage&Partner to become a leading
>software vendor. We also held many meetings explaining our strategy and
>goals to make the PowerUp a useful platform not only for a new
>generation of high-performance applications, but also for the
>development towards a new OS or OS version that fullfills the demands of
>the next century. To our surprise and disappointment Haage&Partner has
>mainly used their status to create a replacement for the PowerUp system
>software which is intended to set a standard controlled by
>Haage&Partner, and they even attempt to trash the complete and
>comprehensive software development in which we have spend many efforts.
Sort of like what happened with your falling out with proDAD also huh?
>At the same time, they have yet not been able to present a stable
>running PowerPC application which the Amiga/PowerUp user would have an
>advantage from.
This is really yet to be seen.
> We highly disregard this contraproductive and
>confrontational behaviour, and will completely drop any support for
>Haage&Partner as a PowerUp developer as well as for any of their
>products.
Oh really? I would like to run their JAVA implementation on the PowerUP
boards and if I cannot, then I am simply not getting one if the support
that is needed to make their products better on yours is not there.
>At this point, it is important to mention that the software concepts of
>phase 5 digital products have all the time been focused on developments
>that open doors and smooth upgrade paths to future and completely
>revised OS versions, which incorporate functionalities such as memory
>protection and multi-processing (which is not supported by the
>Haage&Partner concept, BTW). This has e.g. also been proposed to Amiga
>International/Gateway2000 in May this year, together with our offer for
So what are you going to do if Amiga Inc. planned to support your PowerUP
and they happened to provide some better stuff and solutions, would you
cut THEM off too? No wonder Petro started talking about them not being
interested in PowerUP.. perhaps there was something more to it then met
the eye.
>------------------------------------------------------------
> Wolf Dietrich, General Manager phase 5 digital products
> Return mail address: w...@gf.phase5.de http://www.phase5.de
>------------------------------------------------------------
So now I know how to get the Wolf out of his sheeps clothing.. actually
start talking about products that are better then theirs or are in direct
competition with theirs and THEN by golly, you will see a reply! If only
your customer support via e-mail and the web etc. was so good.
Your closed system and control over everything attitude stinks. Do you
want to be the Microsoft of the Amiga world? Let's see.. so far it is
proDAD, Village Tronic and now Haage & Partner which has offered you some
competition and look where it got them.. cut off and put down and no
support. I was actually looking forward to PowerUP until these posts,
especially from you but I am not so sure anymore if you want your way to
be the ONLY way and any sign of competition or betterment or actual free
will and free choice on the part of developers and the end user being
locked out, then no thank you.. I will wait for Amiga Inc.'s offerings. If
you are this way now about your PowerUP accelerators then what the hell
will you be like with your A\Box. Forget it, if it means being locked into
ONE way for the rest of my computing life.
P.S. I am posting this publicly before I send it off to you personally via
the e-mail address listed.
____________________________________________________________________________
|C= C=| For | Such giants are these! Great shoulders |CD32 /// |
| C64c |Jay Miner -> bear so many. I stand among them. |A500 /// |
|200 MB HD| RIP | - Dave Haynie | /// |
|Jiffy DOS|_________|___________________________________________|\\\///3000|
|Swiftlink| gn...@plato.ucs.mun.ca or geo...@garfield.cs.mun.ca| \XX/A1000|
|C= C=| http://www.cs.mun.ca/~georgen |Team AMIGA|
|---------|-----------------------------------------------------|----------|
| CMD | Euripides - Never that which is shall die! |Amiga Int.|
|_________|_____________________________________________________|__________|
George Noel
In article <342E7C45...@gf.phase5.de>,
Wolf Dietrich <w...@gf.phase5.de> wrote:>> [...]
>>
>> >Furthermore, phase 5 digital products will not tolerate that an
>> >incompatible software takes over the control of the PowerUp hardware and
>> >prevents the user from running software that has been developed for the
>> >PowerUp boards by phase 5 digital products as well as many other
>> >vendors.
>>
>>
>> This is ironic. Even Microsoft pales beside this kind of business practice.
>
[SNIP]
>phase 5 supplies, as I stated, a solution which consists of hardware and
>software. We provide a HAL on which every developer can integrate any
>solution he likes, while staying compatible. As far as the PowerUp
Well this sounds good.
>boards are concerned, we are the system vendor, and believe that we
>should control the system software that controls the hardware underneath
>the HAL, especially to make sure that all future developments by all the
>developers who support PowerUp remain compatible.
But why? Competition is good.. if someone can provide a genuinely better
software, then let them do it.. it is up to the end user to decide which
one is the best and if they or developers are going to support it. If
their software has compatibility issues, genuine ones, not ones forced on
them by you, then it will be looked down upon anyways.
>If you want to make a comparison with Microsoft, you should question the
Yours does look like one also.
>strategy of Haage&Partner, who - as a software and development system
>vendor - want to take over full control of the hardware that we deliver.
>That means that the kernel, the (only compatible) compiler as well as
>some application software - including their java development - come from
>one source.
This is true and I see them trying to bully the pOS development on the
Amiga as well. But you know? I am interested in some of their products no
matter how I can get it.
> They even claim their software could be a standard for
>future PPC Amigas from Amiga Inc. - provided that Amiga Inc. let
>themselves and their OS development be controlled by the H&P solution,
>what H&P doesn't say.
I wouldn't want that.
>Now why has Haage&Partner spent so many efforts in developing an
>incompatible kernel, which is quite similar to our own system software
>and offers no relevant advantages?
This has yet to be fully determined.
> Could it be that they want to control
>the way any future development will go (and sell the software that is
>mandatory for the development)? Or do you think that instead of
Well, I am sure that is one major reason.
[SNIP]
>Also, your translation was wrong. Ralph Schmidt didn't say "WarpUP may
>be compatible with the current boards, but this will change in a week or
>two" but said "WarpUp may run on the boards shipped so far, but this
>will change in two weeks from now." The difference is small, but
>important: WarpUp HAS NEVER BEEN COMPATIBLE, but runs only because
I was figuring it would break upon the introduction of the different CPUs
in the newer accelerator cards.
>Haage&Partner can currently disable the PowerUp system software. As in
>future revision of PowerUp the PPC-library and other parts of our
>software will be located in the Flash-ROM
So not only will you be forcing the issue, you will be changing the
product line in mid production and selling when it is yet so new? MAN,
that is crap. I hope you provide a free upgrade to the people who
purchased your older boards and that inconsistencies like this do not
continue in the future.
> (a feature that had been
>planned from the beginning regardless of the current haage&partner
>situation),
Yeah.. right.
> WarpUp will not longer run as it is incompatible to our
>solution.
And they won't be able to get around it?
I still don't like this monopolistic control on your companies part.
And this couldn't have happened at a WORSE time. I think your responses
have done more damage to your product then anything H&P have done!
>------------------------------------------------------------
> Wolf Dietrich, General Manager phase 5 digital products
> Return mail address: w...@gf.phase5.de http://www.phase5.de
>------------------------------------------------------------
-=*George*=-
Ernst Berg
RC> From: chia...@cli.di.unipi.it (Rudi Chiarito)
RC> Wolf Dietrich (w...@gf.phase5.de) wrote:
RC> >boards are concerned, we are the system vendor, and believe that we
RC> >should control the system software that controls the hardware
RC> underneath
RC> >the HAL, especially to make sure that all future developments by all
RC> the
RC> >developers who support PowerUp remain compatible.
RC> Commodore was the system vendor of Amiga. They didn't care if someone
RC> else
RC> wrote other system software that controled the hardware, like Minix,
RC> BSD, etc.
RC> If a customer buys an hardware product, I guess he/she has the RIGHT to
RC> choose what to run on it.
RC> >Now why has Haage&Partner spent so many efforts in developing an
RC> >incompatible kernel, which is quite similar to our own system software
RC> >and offers no relevant advantages? Could it be that they want to
RC> control
RC> There ARE advantages. I saw a demo (and I had an interview with H&P for
RC> a
RC> magazine) last week running on both libraries and the performance boost
RC> sported
RC> by WarpOS was noticeable. And that was on a mere 603.
Another thing is the question of a PowerPC standard that ALL developers
can use.
I've read on the Picasso mail list that they have an interface that is
400 times faster then the Phase 5 one and that it REPLACES and DOES exactly
what the Phase 5 one does.
So what my point is that who has the standard does not matter. Who
builds a PowerPC card does not matter. But when I buy a PowerPC card I want it
to work with PowerPC software.
So I ask all peoples who are developing something as important as an
accelerator such as the PowerPC to openly adopt standards through peaceful
coexistence with other companies. It is the open competition that creates
invention.
My view is that the Amiga has made it this far because we have ( for
the most part ) worked on a basic level of the need to survive; the Users, the
resellers and the developers.
I am an end user. I own an A2000 and I am still spending to upgrade so my
voice counts.
Thank you
Ernst Berg
* Q-Blue 1.0
--
| Fidonet: Ernst Berg 1:382/311
|Internet: Ernst...@311.ima.infomail.com
|
|Standard disclaimer: The views of this user are strictly their own.
Mans Engman
On Tue, 30 Sep 1997 01:19:20 +0200, Holger Metz <hm...@rz.uni-frankfurt.de>
wrote:
>Mans Engman schrieb:
>then there will be a phase5 sticker on every software :) phase5 approved
>! :)
yep. :)
>hmm, the best way would be for phase5 and H&P that they come together
>for one solution ! Awar would be the last this small amiga community
>would help !
That would definitely help, unfortunately it doesn't seem like
they get along very well... ;/
>I think the best way would be the user/developer jugdes what is best for
>you and installs "his"
>software he wants !!
Yes, but on the other hand I don't want to see a fragmentation right
now. Run P5 software -> reboot -> run warpos software -> reboot...
[snip]
>they give a clue !! you have includes for all assembler and c compiler
>!! THE MAIN problem isthat there are only 2 compilers which produce PPC
>.. Storm C and GCC .. ! Does this give you a clue..?
Well, the only clue I have is that if you want to make WarpOS software
you _need_ to use StormC/StormAsm/StormWhatever, because gcc doesn't
make the "right" kind of executables.
>I think the main reason why phase5 would like all developers stick to
>their card is, that if they dont, they would loose the PPC Market..
>Warpos doesnt stick to the Hardware.. (they do something similar as the
>ISDN Cards on PeCe. One fossil driver for the
>card itself on the Hardwareside and one software (C)API for the
>developers..) so when everyone uses this WarpOS System, other companys
>could produce PowerPC cards which runs with every software produced for
>Warpos!! But if phase5 does something against it.. no other company
>could produce powerpc cards, because they had to be compatible with the
>P5 cards.. (Ralph Schmidt library runs only on phase5 cards) or the
>software you bought (which uses p5 ppc.library) wont run... its the same
>as microsoft does with their visual products.. you can only produce
>software which runs under Microsoft OS. But looking on the other side of
>H&P its nearly the same... when some company wants to produce a powerpc
>card they have to licence warpup from H&P (or write a driver).
Uhh, stick to the hardware what? Can I run WarpOS on that Mac over
there? No. Can I run the p5 ppc.library on that Parsytec GC? No.
ppc.library isn't any more tied to the specific hardware than warpos
is. Right now both run only on p5 cards, so what? Nothing prevents
someone (p5 or anyone else) from writing a ppc.library compatible with
other ppc cards (if/when they arrive, that is ;).
>So to come to a end... the best way it would be to let the user decide
>what software he uses.. or what would YOU
>do if you buy a pc and get told that you should ONLY INSTALL windows95
>on it or the Hardware wont work ??
Well, if you buy a machine from M$, I would expect it to run Windoze. :)
Of course the user should be able to run whatever OS he likes. But now
the main problem is to smoothly integrate the ppc board into the amiga
environment. Two incompatible solutions with similar functionality is
a bad thing.
/Mans
mo...@lysator.liu.se
Wolf Dietrich
This is a multi-part message in MIME format.
--------------12A379C23B2483F135A89F0C
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
George Noel wrote:
>
> > As it does not use the HAL that is provided by
> >phase 5 digital products, but hacks the hardware by using an incomplete
> >knowledge of the hardware functionality - which has been aquired by
> >reengineering the phase 5 digital products software - it is most likely
> >that it will not work with future revision changes or different product
> >releases of the PowerUp series.
>
> This was what I was worried about in the first place when it came to
> PowerUP but it was pointed out, that PhaseV made sure things will work on
> the various CPUs and not be specific to only certain ones. So yes, if this
> is the case, it is a major reason to stay away from WarpOS.
>
> >Furthermore, phase 5 digital products will not tolerate that an
> >incompatible software takes over the control of the PowerUp hardware and
> >prevents the user from running software that has been developed for the
> >PowerUp boards by phase 5 digital products as well as many other
> >vendors.
>
> Now THIS is scary. I do not like that attitude.
You really think this is scary? All that we want to do is to insure that
on a PowerUp board, the software developed by us or by others using the
different available development packages can run AT ANY TIME. We do not
want to see our users getting frustrated because they have to make a
plan which software they may run in parallel or not. We spend work and
money in free software and services such as CyberGL native and the MPEG
stuff (not to mention our comprehensive PowerUp system software), we
provide a free development system, and yes, we also gave out certain
free developer hardware. Isn't it really scary that H&P follows a path
that disables all the results that come out of these efforts? I believe
so. We are not against compatible alternatives, and we supported H&P and
their efforts all the time while they always promised us their stuff
would be completely compatible with ours - until, surprise surprise,
last week.
> > All software which is supplied by phase 5, and in which we have
> >invested lots of work and financial efforts to make the PowerUp a useful
> >product, would be disabled once the Haage&Partner software is installed
> >in the system.
>
> Sure.. so what? A reboot could send it back if it is done properly.
??? Do you really think that's a solution? I can already see the flame
and hate postings from users who would have to reset their system when
they want to use a PowerUp compliant program after they have started a
WarpUp-based program before.
> >would not run. USAGE OF THE THE HAAGE&PARTNER SOLUTION WOULD RESULT IN A
> >COMPLETE INCOMPATIBILITY MESS.
>
> But could provide some competition, no?
What do you mean? At this moment, there is no other PPC hardware. If
someone would come up with such a hardware, it would make sense that we
negotiate and discuss a platform-independent compatibility standard -
and we would be ready for that.
Software-wise the competition argument is useless, as
> > In the interest of our customers, who buy
> >our products with high expectations, and all our developers who expect a
> >completely compatible product line, we will take care that the phase 5
> >system software can not be disabled.
>
> Again, that is scary. And not wanted.
You may let users of the PowerUp boards speak for themselves, if they
want a software that disables many of the applications and the utility
and library software that they got with their boards or buy from other
vendors.
> >further. Additonally, soon an update for SAS/C on the Amiga can be
> >expected, which supports PowerUp and the ELF format.
>
> Hey.. this is great news!
As we appreciate freedom of choice, we are happy to see these efforts
and will support them as good as we can.
> >Haage&Partner, instead, wants to establish their own "extended
> >Hunk-format", which may is derived from the Amiga Hunk format, but still
> >a new and proprietary solution. It is a concept that forces developers
> >to use the only solution supporting this format - the Storm C compiler
> >by Haage&Partner.
>
> But instead, you also want developers to ONLY use your solution so what
> makes you any better in that regard?
We support any developer who wants to build a compiler (see above) or
anything else (which is compatible). Looks like there will soon be two
free products on the market. The H&P compiler costs a lot of money, and
people may be forced to buy it to develop under WarpUp. We do not want
developers to use our solution because of a short-term economical
interest, but I must assume that H&P does want that. And we are not in a
conflict as we provide the PowerUp System Software as a part of our
product solution, and do not sell own software products. You may ask
other software vendors if they see a problem in the H&P approach.
> > Haage&Partner declares it's product being open and
> >Amiga-compliant - in fact, it is only open to the Haage&Partner standard
> >and compliant to their own software solutions.
>
> Geez.. so is yours. Point moot.
Not quite right. We have choosen an industry-standard which allows to to
development for PowerUp using tools running on other system, too.
As a matter of fact, I believe this whole mess started in the early
times when we were looking for a powerful PPC compiler to support
PowerUp. We decided to use the ELF binary format because it opens up
many options, such as using powerful compilers available on other
systems or free compiler such as the GNU. H&P *always* wanted to push
their "extended-Hunk-Format" and make the Storm C the "official" and
only PowerUp development tool - but of course not fro free...
> >As the solution from Haage&Partner is not compatible and can - because
>
> Well so far it is your word against theirs.
??? H&P told us that it is not compatible with our system software
solution.
> >of it's proprietary software concept - not fullfill demands of
> >professional developers,
>
> I say let them speak!
>
> > it is very unlikely that developers will choose
> >it.
>
> Again, let them speak for themselves.
Do we hinder them to speak? We just bring our arguments, as H&P does.
And we protect those developers who develop on top of our software
solution by making sure their software runs on our boards. Just as a
side note, imagine how much support trouble can be caused if the H&P
software disables ours for the software vendors as well as us.
> > phase 5 digital products highly recommends not to use this
> >incompatible system,
>
> That is your choice but you cannot enforce it.
We can make sure that compatible software will work on our boards, and
H&P may make sure that their software is compatible or not.
> > as well as the Storm C compiler which - because of
> >the concept - is wasting most of the performance that the PowerUp boards
> >can deliver (this issue will also be adressed in detail later).
>
> I'd be interesting in reading that but for now, that recommendation was
> not needed and actually by the sounds of some of these talks, unless these
> can be proven, it sounds liable for damages to me.
It is a conceptual problem and covers issues such as programming style.
Please see the attachment below, which is a copy of a posting to this
newsgroup from yesterday.
> But as my personal opinion about what I know about StormC and SAS/C, I'd
> got with SAS/C anyday.
Probably a good choice, my personal opinion.
> >phase 5 digital products has supported Haage&partner with free developer
> >systems, as well as all necessary information to develop compatible
> >software and tools, thus enabling Haage&Partner to become a leading
> >software vendor. We also held many meetings explaining our strategy and
> >goals to make the PowerUp a useful platform not only for a new
> >generation of high-performance applications, but also for the
> >development towards a new OS or OS version that fullfills the demands of
> >the next century. To our surprise and disappointment Haage&Partner has
> >mainly used their status to create a replacement for the PowerUp system
> >software which is intended to set a standard controlled by
> >Haage&Partner, and they even attempt to trash the complete and
> >comprehensive software development in which we have spend many efforts.
>
> Sort of like what happened with your falling out with proDAD also huh?
?? Which falling out? There's a similar situation: ProDAD always wanted
to convince us that their product is so great and we should adopt it as
a standard, as H&P did in the past nine months. Well, we couldn't see
the advantages of their concept (and in my personal opinion, their pOS
is - even on 68k - still far from being a complete OS rather than a
replacement for some GUI and middleware parts of the Amiga OS)
> >At the same time, they have yet not been able to present a stable
> >running PowerPC application which the Amiga/PowerUp user would have an
> >advantage from.
>
> This is really yet to be seen.
So far they could not present e.g. a running PPC version of their
ArtEffects software under their WarpUp kernel - and I thought they
sinply need to recompile the code with Storm C...
> > We highly disregard this contraproductive and
> >confrontational behaviour, and will completely drop any support for
> >Haage&Partner as a PowerUp developer as well as for any of their
> >products.
>
> Oh really? I would like to run their JAVA implementation on the PowerUP
> boards and if I cannot, then I am simply not getting one if the support
> that is needed to make their products better on yours is not there.
Then you should ask them to make their JAVA compatible with our PowerUp
system software, so that their JAVA and e.g. a VRML plugin for a Web
Browser which uses our PPC-native CyberGL library would work together.
What would YOU say if these two products would only work exclusively?
Whom would you blame?
> >At this point, it is important to mention that the software concepts of
> >phase 5 digital products have all the time been focused on developments
> >that open doors and smooth upgrade paths to future and completely
> >revised OS versions, which incorporate functionalities such as memory
> >protection and multi-processing (which is not supported by the
> >Haage&Partner concept, BTW). This has e.g. also been proposed to Amiga
> >International/Gateway2000 in May this year, together with our offer for
>
> So what are you going to do if Amiga Inc. planned to support your PowerUP
> and they happened to provide some better stuff and solutions, would you
> cut THEM off too? No wonder Petro started talking about them not being
> interested in PowerUP.. perhaps there was something more to it then met
> the eye.
Amiga Inc. is or will be the system vendor of their own solution. If
they'd be interested in using technology from us, we are happy to
cooperate with them. if Amiga Inc. decides to port their OS to the
PowerPC and update it, we would for sure apply for an OS license for our
current and future developments. But this is a completely different
issue, I wouldn't compare Amiga Inc. with H&P.
> So now I know how to get the Wolf out of his sheeps clothing.. actually
> start talking about products that are better then theirs or are in direct
> competition with theirs and THEN by golly, you will see a reply! If only
> your customer support via e-mail and the web etc. was so good.
Nice comparison ;-) But we are not reacting on a competition in a
commercial meaning of the word, but only on an attempt to take over the
control over a hardware platform, which is driven by strong commcercial
interests of H&P.
> Your closed system and control over everything attitude stinks.
What do you mean? We have developed a very complex hardware concept, and
the first product of this new line is out now. Other products will
follow, which may offer different options. To make sure that software
will be compatible, we provide a HAL. From that level on, it is an open
system. Software such as WarpUp that hacks the hardware may not be
compatible with other baord version when there are new developments and
hardware or revision changes. And the hardware design of our own
products WILL remain under our control.
As a matter of fact, someone must control a system (well, not really
someone but somesoftware ;-)) Why should it be the software that has
been hacked by a third party, who has reengineered our software, does
not know about the full functionality and options of the hardware, and
who only claims his software would be better than ours in a massive
marketing campaign, which very directly indicates that our software
would not be sufficient?
> Do you
> want to be the Microsoft of the Amiga world?
No, but I definitely believe H&P want to be that, and their strategy
look like they had their training lessons there.
> Let's see.. so far it is
> proDAD, Village Tronic and now Haage & Partner which has offered you some
> competition and look where it got them.. cut off and put down and no
> support.
??? As a matter of fact, some of these companies seemed not to be
interested in our support or readyness for cooperation, because they all
followed own projects which they have given higher priority, what is
their own decision. It was e.g. ProDADs decision to focus on their pOS
and obviously not follow their first announcements to support PowerUp
with ported version of their software. Well, so we have a 68k pOS
prerelease now, but no PowerUp Adorage or other useful product from
them. H&P had and appreciated our support, and now consequently are
trying to cut us and our system software development off.
> I was actually looking forward to PowerUP until these posts,
> especially from you but I am not so sure anymore if you want your way to
> be the ONLY way and any sign of competition or betterment or actual free
> will and free choice on the part of developers and the end user being
> locked out, then no thank you.. I will wait for Amiga Inc.'s offerings. If
> you are this way now about your PowerUP accelerators then what the hell
> will you be like with your A\Box. Forget it, if it means being locked into
> ONE way for the rest of my computing life.
It is your choice which system you may prefer, but I don't think it is
fair (do you say "politically correct" in the US) to say that we would
lock out free will and the freedom of choice while we just want to
prevent us from being locked out from the control/influence on the
development of and on our own product.
Also, again I have to reinforce that we can see no advantages of the H&P
software. It is of course promoted by them as being superior, but they
will have to proove, and we are sure that our solutions provides a
better concept and performance.
------------------------------------------------------------
Wolf Dietrich, General Manager phase 5 digital products
Return mail address: w...@gf.phase5.de http://www.phase5.de
------------------------------------------------------------
--------------12A379C23B2483F135A89F0C
Content-Type: text/plain; charset=us-ascii; name="h&p-argue3.asc"
Content-Transfer-Encoding: 7bit
Content-Disposition: inline; filename="h&p-argue3.asc"
>From - Mon Sep 29 16:36:39 1997
Path: news.nacamar.de!not-for-mail
From: Wolf Dietrich <w...@gf.phase5.de>
Newsgroups: comp.sys.amiga.misc
Subject: Re: WarpOS released !!!
Date: Mon, 29 Sep 1997 16:33:05 +0200
Organization: phase 5 digital products
Lines: 73
Message-ID: <342FBC21...@gf.phase5.de>
References: <0200020622...@BIRDLAND.es.bawue.de> <342E63DF...@gf.phase5.de> <60lu0a$a...@arcadia.informatik.uni-muenchen.de> <342E7C45...@gf.phase5.de> <60ntac$jhi$8...@pania.unipi.it> <342F7A23...@gf.phase5.de> <60o4aj$jhi$1...@pania.unipi.it>
NNTP-Posting-Host: ufoo.phase5.de
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
X-Mailer: Mozilla 3.0Gold (X11; I; Linux 2.0.0 i586)
Xref: news.nacamar.de comp.sys.amiga.misc:178290
------------------------------------------------------------
Wolf Dietrich, General Manager phase 5 digital products
Return mail address: w...@gf.phase5.de http://www.phase5.de
------------------------------------------------------------
--------------12A379C23B2483F135A89F0C--
Rudi Chiarito
I forgot to reply to the following...
Wolf Dietrich (w...@gf.phase5.de) wrote:
>future revision of PowerUp the PPC-library and other parts of our
>software will be located in the Flash-ROM (a feature that had been
>planned from the beginning regardless of the current haage&partner
>situation), WarpUp will not longer run as it is incompatible to our
>solution.
Do you consider that a major problem? Nothing is going to prevent users from
"upgrading" their Flash-ROMs with the SW of their choice. Be it from Phase5,
Haage&Partner or anyone else. A plain and simple solution :)
When the user has bought your hardware, he/she can do whatever he/she likes
with it. He can even play cricket with the board: you can't do anything to
prevent that from happening.
Rudi Chiarito
Wolf Dietrich (w...@gf.phase5.de) wrote:
>H&P have been promoting their Storm C with a concept that you simply
>recompile the existing C code which has been written for 68k Amiga and
>tell the compiler to generate PPC code. This may be very comfortable for
That's ONE way to get things done. It isnt the only one though.
>the developer, but is completely unefficient. Not only that in the worst
>case you have a random mix of 68k and PPC code, but also the two CPUs
AFAIK with StormC the programmer has control over which parts use PPC code
and which don't...
>will not run in parallel as it is one continuous program flow which
>jumps between PPC and 68k. If a linear code sequence is executed, the
>68k and the PPC always work one after the other, in the worst case with
>lots of switches and a useless overhead.
>A significant loss of performance is the result of this way of porting
>applications to the PowerPC. It could result in a scenario where
>developers use Storm C, recompile their code, find that it is two times
>faster than on a 68060 and decide that this is OK, while with some
>thinking and conceptual consideration the program may would work five or
>ten times faster than on the 68060. Insofar this is a poor solution from
>which maybe a less motivated developer would benefit, being enabled to
>do a quick-and-dirty conversion, but not the customer, who would get
>products that do not exploit the performance of his hardware at all.
That makes no sense at all for me. A bad programmer is a bad programmer, no
matter what tools he's using. If he's too lazy to rewrite the program to
take full advantage of PowerUps, you can be sure he will write bad programs
even if he's using your tools. StormC's approach makes your life easier, but
it doesnt enforce you to adopt a lazy strategy.
And a good programmer is always a good programmer: he won't be satisfied if
a simple recompile gives him a program which is two times faster. He will
always try to get the most out of his hardware.
What you're talking about would be the developer's fault, not the
development tool's.
Rudi Chiarito
Wolf Dietrich (w...@gf.phase5.de) wrote:
>Yes, and the demo program running under the PowerUp System Software has
>also been provided by them. Oh, by the way, was it the "Set Pixel"-Demo
>which they have shown you? If so, this demo is as useful as testing a
>Porsche by driving backwards.
I'm sorry, but actually it was the 'voxelspace' demo. Far from being a
useless one. I also saw their Mandelbrot generator (with realtime zooming
and panning, just like the good old Mand2000). And they didn't show me any
"Set Pixel" demo.
BTW, I wonder why H&P have been promoting the PowerUP for you.. I mean,
they were both last year in Milan at IPISA and this year in Empoli (Florence)
at Pianeta Amiga, showing off *your* boards.
Phase5 wasn't there, both times. Why?
Mans Engman
On Tue, 30 Sep 1997 10:21:47 GMT, Mathias Ortmann
<ort...@informatik.tu-muenchen.de> wrote:
>mo...@bulldozer.rydnet.lysator.liu.se (Mans Engman) wrote:
>>Yes, but on the other hand I don't want to see a fragmentation right
>>now. Run P5 software -> reboot -> run warpos software -> reboot...
>Why is frequent rebooting only a problem when it happens on the powerUP?
Having to reboot is always a problem. I haven't said anything else.
>When the PIOS ONE was still an issue, nobody complained about having to
>reboot to switch between applications available for MacOS only, Linux-
>specific programs and BeOS-specific software (is there any?). Worst of all,
I'm not too involved in the PIOS concept. I don't care and don't complain.
My interests lie in extending the Amiga functionality. (Huh, real amiga
freak... ;)
>partition accessibility will be low (can a Mac read/write ext2fs? does
>Linux have a BeOS fs yet?) - the multi-OS approach and the concept of
>multitasking seem to be mutually exclusive.
Yep, that's a problem. That's why it's pointless to launch a new "OS"
which needs proprietary tools and is incompatible with a similar system.
But that's MHO of course.
Gene Heskett
Gene Heskett sends Greetings to George Noel;
Refering to a message from Wolf D.
>I still don't like this monopolistic control on your companies part.
>And this couldn't have happened at a WORSE time. I think your responses
>have done more damage to your product then anything H&P have done!
George, if Wolf's company wants to design something that should be
operated in a specified manner for the best overall results, then he
can _do_ it. I don't think for a minute that he did it out of spite
for H&P. The specs are I think, pretty well given. H&P didn't choose
to use them, then thats their problem.
However, lets put this in the proper perspective. Will your new car
work equally well if instead of 87 octane, you put some diesel fuel
of a reasonable cetane number it there instead? Note that I said
'work equally well' there. Yes, it will run, but not quietly, nor
with much power, prompting you to either run it on through, driving
very gently, and diluting it with the best gasolene sold everytime
theres room for another gallon in the tank, or have the tank drained.
Been there, done that, tanker filled wrong tank with super-diesel!
If Wolf wants it to run his grade of fuel, and you don't agree about
that attitude, then either get over it, or go buy somebody elses.
What, nobody else is making this model? Then get over it.
Cheers, Gene
--
Gene Heskett, CET, UHK | Amiga A2k/2630/2632 56 megs fast/2 megs chip
Ch. Eng. @ WDTV-5 | A2091/GuruRom-1.08 gig, CDROM, Multiface III
<gene_h...@iolinc.net> or | AlfaData2008-IDE + 345meg & 2.2 gig drives
<gene_h...@westvirginia.net>| 525 meg tape, Stylus Pro, EnPrint
This space for rent, cheap | Picasso-II, C=1960
--
Carl Skedinger
George Noel wrote:
>
> In article <1200010623...@BIRDLAND.es.bawue.de>,
> Steffen Haeuser <mag...@birdland.es.bawue.de> wrote:
>
> >Of course, as the coder of ppc.library outlined, it is not completely the
> >fault of the software... 68k/PPC switch is a hardware problem, also. But
>
> claims the boards are fully multiprocessing and parallel.
>
George, if there is something you don't understand,
try not to post, ok?
/Calle
--
------------------------------------------------------------------------
Calle Skedinger e92...@e.kth.se
Jason Compton
Rudi Chiarito (chia...@cli.di.unipi.it) wrote:
: BTW, I wonder why H&P have been promoting the PowerUP for you.. I mean,
: they were both last year in Milan at IPISA and this year in Empoli (Florence)
: at Pianeta Amiga, showing off *your* boards.
: Phase5 wasn't there, both times. Why?
Wolf doesn't always come to my parties when I invite him, either. :)
Seriously, your example is hardly damning. From a strictly objective
viewpoint, H&P wants to sell their PowerPC software. If they're going to
be able to do that, they need people to buy PowerUP boards. Simple.
--
Jason Compton jcom...@xnet.com
Editor-in-Chief, Amiga Report Magazine Anchor, Amiga Legacy
http://www.cucug.org/ar/ http://www.xnet.com/~jcompton/
There are only dreams... ...like any other.
Mathias Ortmann
d9...@indy.net wrote:
> > Why is frequent rebooting only a problem when it happens on the powerUP?
> >
> > When the PIOS ONE was still an issue, nobody complained about having to
> > reboot to switch between applications available for MacOS only, Linux-
> > specific programs and BeOS-specific software (is there any?). Worst of all,
> > partition accessibility will be low (can a Mac read/write ext2fs? does
> > Linux have a BeOS fs yet?) - the multi-OS approach and the concept of
> > multitasking seem to be mutually exclusive.
>
>Boot Shapeshifter and learn...
Shapeshifter would have required UAE or something similar underneath it
to run on the PIOS ONE...
--
Mathias Ortmann ort...@informatik.tu-muenchen.de
UAE 0.6.9 Win32/DirectX @ http://www.informatik.tu-muenchen.de/~ortmann/uae/
Milan Polle
Hello? What is going on here, are we taking the side of H&P?
Not me, I think what they have don't is totally useless, stupid and
a bad thing for the Amiga.
They are fragmenting the Amiga community even more with their obsolete
OS. You won't gain anything from using their system, they too have the
delay when switching processors (read the docs).
Besides that, the tone of their guide explaining how to make games
not only sounds patronizing, but it also tells you that you have to
use their system for making games or else it won't work well.
Then they are telling you how fantastic Steffan Haussers system is
for making PPC games, has it already been tested with PPC or are they
just trying to get him on their side as a good advocate ( I don't mean
this negatively about Steffan, I like the guy, but stay on P5's side).
Why in the world would they make a system that is equal to P5's one?
I think it is because they would like to have a major part in the making
of the upcoming official Amiga OS.
What a waste of time, energy and probably good programmers, they should
have done something positive for the Amiga.
Storm C doesn't have a very good reputation either, I would rather use
SAS/C PPC. And as they don't have complete knowledge about the PowerUp,
how could their system be better?
Phase5 has the right to develop their own 'driver' for their card, if
WarpOS would mess up compatability between programs (how would you like
to switch between normal PowerUp programs and WarpOS ones by rebooting
all the time?) they would mess up the reputation of PowerUp and it's
success. P5 doesn't forbid developing systems for their board (like
Linux) but they have to be based on their core, for compatability
reasons.
CBM's own products would most of the time be the most compatible, using
their own software, if you remember.
What a shame, I like H&P for their ArtEffect effort, I want them to do
positive things for the Amiga, not make things more difficult.
And the bloated advertising with giving their system away for free is
just the way Microsoft got where they are now.
Lastly, did anybody read P5's statement about POs? They where right,
the PowerUp boards are here (very delayed) and where is POs PPC or
PowerUp?
It isn't there.
Please people, don't be so hard on Phase5, they are doing great things
for the Amiga, we have their president responding to all your rantings
for crying out loud! This has got to stop.
Milan
------------------------------------------------------------------------
Imagine textures/images/utils, CandyFactory instant 3D graphics program,
Atari Lynx Programming (with Amiga), vic-20 emulator etc.
All at: http://www.geocities.com/SiliconValley/Way/1038/
------------------------------------------------------------------------
George Noel
In article <1200010623...@BIRDLAND.es.bawue.de>,
Steffen Haeuser <mag...@birdland.es.bawue.de> wrote:
[SNIP]
>Of course, as the coder of ppc.library outlined, it is not completely the
>fault of the software... 68k/PPC switch is a hardware problem, also. But
Funny.. you wouldn't say from one of Wolf Dietrich's recent posts.. he
claims the boards are fully multiprocessing and parallel.
[SNIP]
>Steffen Haeuser
-=*George*=-
Steffen Haeuser
ĸ° chia...@cli.di.unipi.it schrieb : ĸī
Hi!
ch> There ARE advantages. I saw a demo (and I had an interview with H&P for a
ch> magazine) last week running on both libraries and the performance boost sported
ch> by WarpOS was noticeable. And that was on a mere 603.
I heard there will be still an update of WarpOS which will speed up it even
more (not sure, if it is already included into the package on the
haage&partner site...)
Steffen Haeuser
Espen Berntsen
On 29 Sep 1997 22:20:22 GMT, mo...@bulldozer.rydnet.lysator.liu.se
(Mans Engman) wrote:
>>Say what? A new bogged down hardware key to prevent other from making
>>things for the new hardware? Like M$ did for NT when they used
>>undocumented lib calls to make email programs (and thus preventing
>>many other vendors".
>
>Ehh, I don't see an analogy. An analogy would be that someone develops
>a Windoze application, that when run shuts down all other Windoze
>programs, and M$ refuses to give this program a "M$-Windoze-approved"
>sticker.
No, they add some OSkey thath prevents that program from executing.
>>So if it finds some H&P stuff you disable what you can? How nice.
>>Remind me of not getting the alleged A\Box.
>
>No, no, no. Running WarpOS means that P5 software, or software created
>with P5's _free_ devkit won't run anymore. Such a solution is a big step
>backwards, back to the good old days when you had to reset the computer
>between running programs.
I agree that the H&P approach is bad when it mucks around with the
system. And I am not impressed with the StormC and will stick with
SAS/C. Still I do not think a hardware key to prevent a lib from
working is the right way to go.
>On the other hand I should note that WarpUP contains two versions
>of powerpc.library -- one that seems to be compatible with ppc.library
>(a layer "on top" of it), and one version using WarpOS which will
>grab the ppc board completely. So programs can be made to work for
>both cases. But any program that needs WarpOS-specific features needs
>the right library to run, and will thus disable the possibility to run
>ELF binaries at the same time. IMHO, stupid.
I got the impression that the WarpOS would be used for games mostly?
If so, the multitasking ain't too much of an issue. If it is a general
lib used for apps it certainly IS an issue. Still it doesn't change my
view that if P5 will add a hardware key to prevent warpOS from running
(or programs using the WarpOS) they are doing it just for spite and it
is yet another reason not to use P5 products.
>>And the PowerUp board is only compliant with P5 stuff. And the big
>>difference here is???
>
>Developing the P5 way means; both free compilers and tools (gcc etc),
>and also (possibly) the choice to use other industry-standard tools that
>can work with ELF objects. Now tell me, what way other than the commercial
>Storm-tools is there to compile his/her applications to the H&P format?
>The WarpUp docs don't give a clue, anyhow.
As stated I do not like either of the companies approach. They
obviously are completely unable to co-exist. I hope it backfires. On
both of them.
Wolf Dietrich
Steffen Haeuser wrote:
> Well, i understand your point... but if this is your main interest, why did
> you decline all cooperation with Haage&Partner ? As i was told, they offered
> this once more to your firm, before they actually released WarpOS... in fact
> they even did not leak ANY information about it, before they spoke to Phase 5
Do you think that the fact that we supplied free developer boards and
all kind of support to them indicates that we "declined all
cooperation"? They did not offer support, but frequently asked us to use
their binary format and make Storm C the only and official development
tool.
In fact, they did develop their software and their strategy over a long
time, and promised all the time that their solution would be compatible
with the PowrUp system software. On the 25th of September they for the
first time informed us about their WarpUp stuff, and started a massive
marketing campaign the next day.
> ... i think the HAL to be used with PowerUP should be:
> - efficient
> - comfortable in usage
> - support modern compilers
That's what our HAL does. It even does support compilers from different
vendors and on different platforms.
> I think this should be more important than who does the software... why not
> shipping WarpOS with PowerUP ? Then you would not have to call it "a foreign
> solution". I am sure, Haage&Partner would not have a problem with the
> permission to use it for free.
No, and they would have even less problems to see all developers or
users who want to write software buying their Storm C or PowerASM as no
other solution is available.
> I think it is no use, if Haage&Partner and Phase 5 say things like that to
> each other. I read your statement as to Haage&Partner wanting to overtake the
> market, i read statements from people at Haage&Partner that Phase 5 would want
> to overtake the market and would not deal good with concurrence situations.
I am sorry that the situation forces me to speak very frankly, but I
don't see reasons to hide the truth in nice words.
> Such statements, i believe have no place in a business situation. If i compare
> Haage&Partner and Phase 5 statements - well, i believe the one of H&P was much
> more fair to the other firm.
You should read the WarpUp announcement as well as the Amiga guides
enclosed with WarpUp again, and reconsider if you believe the
indications and frank comments about the phase 5 software solution are
fair.
> According to the programmer of WarpOS simply because the new kernel is more
> efficient. And no, he says, this is NOT based on Beta Versions and Developper
> Boards. Also the way it is included into the system would be "more OS legal".
> And if it is so similar, why do you at all have problems with it ?
THEY say it is more efficient. THEY claim it to be faster or better. We
deny this. And as it is similar in functionality, why should it take
over the hardware completely and stop software that uses the original
PowerUp System Software from working?
Randy Vice
On Tue 30-Sep-1997 3:39p, e92...@elixir.e.kth.se wrote:
e> George Noel wrote:
e> >
e> > In article <1200010623...@BIRDLAND.es.bawue.de>,
e> > Steffen Haeuser <mag...@birdland.es.bawue.de> wrote:
e> >
e> > >Of course, as the coder of ppc.library outlined, it is not completely
e> the
e> > >fault of the software... 68k/PPC switch is a hardware problem, also.
e> But
e> >
e> > Funny.. you wouldn't say from one of Wolf Dietrich's recent posts.. he
e> > claims the boards are fully multiprocessing and parallel.
e> >
e> George, if there is something you don't understand,
e> try not to post, ok?
;-)
: damo...@nostromo.gate.net : Bruce Morrow,a man before and after his time:
:"The Constitution shall never be construed to prevent the people of the :
:United States who are peaceable citizens from keeping their own arms." :
: - Samuel Adams : Morrow Project Planning: Looking forward to the end! :
: TeamONE listserver -> List...@nostromo.gate.net add name@site TeamONE :
Mathias Ortmann
mo...@bulldozer.rydnet.lysator.liu.se (Mans Engman) wrote:
>Yes, but on the other hand I don't want to see a fragmentation right
>now. Run P5 software -> reboot -> run warpos software -> reboot...
Why is frequent rebooting only a problem when it happens on the powerUP?
When the PIOS ONE was still an issue, nobody complained about having to
reboot to switch between applications available for MacOS only, Linux-
specific programs and BeOS-specific software (is there any?). Worst of all,
partition accessibility will be low (can a Mac read/write ext2fs? does
Linux have a BeOS fs yet?) - the multi-OS approach and the concept of
multitasking seem to be mutually exclusive.
Just wondering...
Goswin Brederlow
Wolf Dietrich <w...@gf.phase5.de> writes:
> George Noel wrote:
> > > As it does not use the HAL that is provided by
> > >phase 5 digital products, but hacks the hardware by using an incomplete
> > >knowledge of the hardware functionality - which has been aquired by
> > >reengineering the phase 5 digital products software - it is most likely
> > >that it will not work with future revision changes or different product
> > >releases of the PowerUp series.
It doesn't use the HAL? I heard something very different. And if P5
stops suporting the HAL....
> > This was what I was worried about in the first place when it came to
> > PowerUP but it was pointed out, that PhaseV made sure things will work on
> > the various CPUs and not be specific to only certain ones. So yes, if this
> > is the case, it is a major reason to stay away from WarpOS.
> > >Furthermore, phase 5 digital products will not tolerate that an
> > >incompatible software takes over the control of the PowerUp hardware and
> > >prevents the user from running software that has been developed for the
> > >PowerUp boards by phase 5 digital products as well as many other
> > >vendors.
> > Now THIS is scary. I do not like that attitude.
Yes this is scary. Will they tolerate Linux, NetBSD or other free
software. Will they throw down anything that might be simmilar to
their software? Do they want a monopoly?
I want to run Unix on my computer and that would certainly not work
with P5 software, it would also be version dependant. It would be
worse than WarpOS. Will they give out information about the hardware
so people can run unix on it? It doesn't look like it.
I think they could gain much more customers by giveing out the specs
than they could possible loose.
I heard from several people that they bought this and this Graphics
card and not CV64-3D because there weren't drivers for Unix.
After asking for specs for the CV64-3D mighself I was told that they
had given out the specs to programmers as well and that there would be
a driver. Half a year later I was so pressing that I got the specs
around p5 back from someone else and 24 h later my Linux was running
on my CV64-3D. Of cause its not a fully funktional driver jet, but I
can use the card and use X to enhance the driver further. What have
the programmers P5 gave the specs done in half a year? Where there
any, I sometimes wonder.
> You really think this is scary? All that we want to do is to insure that
> on a PowerUp board, the software developed by us or by others using the
> different available development packages can run AT ANY TIME. We do not
> want to see our users getting frustrated because they have to make a
> plan which software they may run in parallel or not. We spend work and
> money in free software and services such as CyberGL native and the MPEG
> stuff (not to mention our comprehensive PowerUp system software), we
> provide a free development system, and yes, we also gave out certain
> free developer hardware. Isn't it really scary that H&P follows a path
> that disables all the results that come out of these efforts? I believe
> so. We are not against compatible alternatives, and we supported H&P and
> their efforts all the time while they always promised us their stuff
> would be completely compatible with ours - until, surprise surprise,
> last week.
Look at what made the Amiga so great. It wasn't the lack of comercial
and bussiness software, it wasn't the slower and more expensive
hardware (lately). What made tha Amiga great and made it survive was
the open minded concept. I even got blueprints of my Amiga when I
bough my first one. The Operation system is documented and the
documentation released freely (on Fish-disk for example). The Amiga
was open Hardware.
The Result was that people wrote Demos that didn't run on AGA or
040+. Some Bootloader didn't work anymore.
Some people blamed the newer amiga for it, but most blamed the
programmer of that lame stuff that didn't work.
The great benefit was the amount of Shareware and Free Software and
also many System enhancements.
> > > All software which is supplied by phase 5, and in which we have
> > >invested lots of work and financial efforts to make the PowerUp a useful
> > >product, would be disabled once the Haage&Partner software is installed
> > >in the system.
How do you know that they don't work together? Have you tested it?
Since the H&P soft uses the HAL as well it should be no problem,
shouldn't it?
> > Sure.. so what? A reboot could send it back if it is done properly.
> ??? Do you really think that's a solution? I can already see the flame
> and hate postings from users who would have to reset their system when
> they want to use a PowerUp compliant program after they have started a
> WarpUp-based program before.
Of cause rebooting is no solution, but would you thing they blame P5
or wouldn't they blame WarpOS? Of cause there would be some who blame
god and the world as there always are.
> > >would not run. USAGE OF THE THE HAAGE&PARTNER SOLUTION WOULD RESULT IN A
> > >COMPLETE INCOMPATIBILITY MESS.
> > But could provide some competition, no?
> What do you mean? At this moment, there is no other PPC hardware. If
> someone would come up with such a hardware, it would make sense that we
> negotiate and discuss a platform-independent compatibility standard -
> and we would be ready for that.
> Software-wise the competition argument is useless, as
I don't think its useless. There are cgfx and P96. Do you think they are
useless? Why stop anyone from writeing a better driver/device or lib
for something. The New Soft wont start working without the hardware
and if its incompatible users wont use it.
I would like to have protected memory and virtual memory for my
programms. From the doku and some talks on irc I think WarpOS can be
made to provide that whereas the ppc.lib can't. That would make big
difference. If thats true, which I can only test and see when I have
doku for the ppc.lib, it would be a major point in decideing what to
use.
> > > In the interest of our customers, who buy
> > >our products with high expectations, and all our developers who expect a
> > >completely compatible product line, we will take care that the phase 5
> > >system software can not be disabled.
> > Again, that is scary. And not wanted.
> You may let users of the PowerUp boards speak for themselves, if they
> want a software that disables many of the applications and the utility
> and library software that they got with their boards or buy from other
> vendors.
I will want to disable everything and start unix. AmigaOS has no
Memory protection, no virtual memory, no real network support and
other think only unix can give at the moment.
Under AmigaOS I would also like to add memory protection to my
programms and virtuall memory. That would need to take over some
funktions from the ppc.lib and drive some hardware directly.
I think many users would prefer to use memoryprotection, even if its a
hack. Building it without proper doku and support will only make it
more difficult to implement and more buggy.
> > >further. Additonally, soon an update for SAS/C on the Amiga can be
> > >expected, which supports PowerUp and the ELF format.
> > Hey.. this is great news!
> As we appreciate freedom of choice, we are happy to see these efforts
> and will support them as good as we can.
Is this only if the software doesn't relate to a similar p5 software
or a more general statement?
> > >Haage&Partner, instead, wants to establish their own "extended
> > >Hunk-format", which may is derived from the Amiga Hunk format, but still
> > >a new and proprietary solution. It is a concept that forces developers
> > >to use the only solution supporting this format - the Storm C compiler
> > >by Haage&Partner.
> > But instead, you also want developers to ONLY use your solution so what
> > makes you any better in that regard?
> We support any developer who wants to build a compiler (see above) or
> anything else (which is compatible). Looks like there will soon be two
> free products on the market. The H&P compiler costs a lot of money, and
> people may be forced to buy it to develop under WarpUp. We do not want
> developers to use our solution because of a short-term economical
> interest, but I must assume that H&P does want that. And we are not in a
> conflict as we provide the PowerUp System Software as a part of our
> product solution, and do not sell own software products. You may ask
> other software vendors if they see a problem in the H&P approach.
I dont really care about the hunk format unless I have to hack around
in it. If H&Ps solution works, fine. It has some features which are
nice, like mixed binaries. On the other hand I like the p5 decision to
take on a better format (it was elf, wasn't it?) which is widely
spread. I also use gcc which will be provided with the Board, so I
will use p5 format with the gcc, unless I find a good reason to
change.
> > > Haage&Partner declares it's product being open and
> > >Amiga-compliant - in fact, it is only open to the Haage&Partner standard
> > >and compliant to their own software solutions.
> > Geez.. so is yours. Point moot.
> Not quite right. We have choosen an industry-standard which allows to to
> development for PowerUp using tools running on other system, too.
> As a matter of fact, I believe this whole mess started in the early
> times when we were looking for a powerful PPC compiler to support
> PowerUp. We decided to use the ELF binary format because it opens up
> many options, such as using powerful compilers available on other
> systems or free compiler such as the GNU. H&P *always* wanted to push
> their "extended-Hunk-Format" and make the Storm C the "official" and
> only PowerUp development tool - but of course not fro free...
Good choise made. Selling Hardware without Software to run on is
difficult, selling it without a compiler to compile Software for it is
impossible.
> > >As the solution from Haage&Partner is not compatible and can - because
> > Well so far it is your word against theirs.
> ??? H&P told us that it is not compatible with our system software
> solution.
Can anyone test that? (Anyone not already in favour in one or the other)
> > >of it's proprietary software concept - not fullfill demands of
> > >professional developers,
> > I say let them speak!
> > > it is very unlikely that developers will choose
> > >it.
> > Again, let them speak for themselves.
> Do we hinder them to speak? We just bring our arguments, as H&P does.
> And we protect those developers who develop on top of our software
> solution by making sure their software runs on our boards. Just as a
> side note, imagine how much support trouble can be caused if the H&P
> software disables ours for the software vendors as well as us.
If the H&P software does make so much problems, give the suport the
number of H&P support and tell them to ask first if the user runs H&P
software at the same time.
> > > phase 5 digital products highly recommends not to use this
> > >incompatible system,
> > That is your choice but you cannot enforce it.
> We can make sure that compatible software will work on our boards, and
> H&P may make sure that their software is compatible or not.
You can allways change the software slightly to make other software
fail. Some rule will allways be bend. Some rules might be broken
unnoticed. The result of you changeing the soft so those bend rules
break H&Ps programms would result in a change in H&Ps Software so it
works again and the users will be stuck in a constant update war. The
looser will be the user and then both of you will loose.
> > > as well as the Storm C compiler which - because of
> > >the concept - is wasting most of the performance that the PowerUp boards
> > >can deliver (this issue will also be adressed in detail later).
> > I'd be interesting in reading that but for now, that recommendation was
> > not needed and actually by the sounds of some of these talks, unless these
> > can be proven, it sounds liable for damages to me.
> It is a conceptual problem and covers issues such as programming style.
> Please see the attachment below, which is a copy of a posting to this
> newsgroup from yesterday.
Make a black list wich state the programm and the violation. Send
those violations to the programmer and after some time make them
public. The big problem with that is that you have to have that list
uptodate and not 3 Month out of date.
> > But as my personal opinion about what I know about StormC and SAS/C, I'd
> > got with SAS/C anyday.
> Probably a good choice, my personal opinion.
> [snip]
> > > We highly disregard this contraproductive and
> > >confrontational behaviour, and will completely drop any support for
> > >Haage&Partner as a PowerUp developer as well as for any of their
> > >products.
> > Oh really? I would like to run their JAVA implementation on the PowerUP
> > boards and if I cannot, then I am simply not getting one if the support
> > that is needed to make their products better on yours is not there.
> Then you should ask them to make their JAVA compatible with our PowerUp
> system software, so that their JAVA and e.g. a VRML plugin for a Web
> Browser which uses our PPC-native CyberGL library would work together.
> What would YOU say if these two products would only work exclusively?
> Whom would you blame?
Take up the idea about the black list. Users will drive Programms with
violations bankrupt.
> > So now I know how to get the Wolf out of his sheeps clothing.. actually
> > start talking about products that are better then theirs or are in direct
> > competition with theirs and THEN by golly, you will see a reply! If only
> > your customer support via e-mail and the web etc. was so good.
> Nice comparison ;-) But we are not reacting on a competition in a
> commercial meaning of the word, but only on an attempt to take over the
> control over a hardware platform, which is driven by strong commcercial
> interests of H&P.
Have you seen the adds for PowerUP place in the last AmigaMagazins?
They still say that this and this will be available some month gone by
now. Where are the boards the adds says should be available two month
ago? Same for the webpages. Sometimes there were half a year out of date.
> > Your closed system and control over everything attitude stinks.
> What do you mean? We have developed a very complex hardware concept, and
> the first product of this new line is out now. Other products will
> follow, which may offer different options. To make sure that software
> will be compatible, we provide a HAL. From that level on, it is an open
> system. Software such as WarpUp that hacks the hardware may not be
> compatible with other baord version when there are new developments and
> hardware or revision changes. And the hardware design of our own
> products WILL remain under our control.
> As a matter of fact, someone must control a system (well, not really
> someone but somesoftware ;-)) Why should it be the software that has
> been hacked by a third party, who has reengineered our software, does
> not know about the full functionality and options of the hardware, and
> who only claims his software would be better than ours in a massive
> marketing campaign, which very directly indicates that our software
> would not be sufficient?
The hardware is great, but the suport (better the giveing out of
specs) needs to be a bit better. Why not place the specs in the manual
describeing the hardware. The only think that can haven is some
software that MAY not work on every system. If the software is
incompatible or breaks rules place it on a black list.
In general you won't have enough time to update the black list with a
incompatible programm on it, cause it will be gone before you update
the list. If the programm last long enough for you to update the black
list, the programm is eigther really needed by the user so he will
live with the incompatibility or it is working fine.
> [snip]
> ------------------------------------------------------------
> Wolf Dietrich, General Manager phase 5 digital products
> Return mail address: w...@gf.phase5.de http://www.phase5.de
> ------------------------------------------------------------
May the Source be with you.
Mrvn
PS: I'm looking forward to haveing a PowerUP board. I think it will
be a great speedup for many of my programms. As soon as I have one I
will start porting my programms and I will use what works best and
feels best. Focused also on what I have to develope. (That probably
leaves me with gcc and p5's lib anyway, so don't start fearing
anything).
To Wolf Dietrich: Is your mail account broken? I'm still waiting for
eigther my board or a reply of my mail. Should I send the mail again?
Sorry for using this method, but my mails seem to get lost somewhere
on the way to p5.
Stefan Burstrom
On 30-Sep-97 11:21:47 Mathias Ortmann (ort...@informatik.tu-muenchen.de) wrote:
>mo...@bulldozer.rydnet.lysator.liu.se (Mans Engman) wrote:
>>Yes, but on the other hand I don't want to see a fragmentation right
>>now. Run P5 software -> reboot -> run warpos software -> reboot...
>Why is frequent rebooting only a problem when it happens on the powerUP?
Because I'd like to be able to run what software I want in what combination
I want. I'd be rather angry if I cannot run Art Effect and play mp3's at
the same time.
d9...@indy.net
--
On Tue, 30 Sep 1997 10:21:47 GMT, ort...@informatik.tu-muenchen.de (Mathias Ortmann) wrote ...
> mo...@bulldozer.rydnet.lysator.liu.se (Mans Engman) wrote:
>
> >Yes, but on the other hand I don't want to see a fragmentation right
> >now. Run P5 software -> reboot -> run warpos software -> reboot...
>
> Why is frequent rebooting only a problem when it happens on the powerUP?
>
> reboot to switch between applications available for MacOS only, Linux-
> specific programs and BeOS-specific software (is there any?). Worst of all,
> partition accessibility will be low (can a Mac read/write ext2fs? does
> Linux have a BeOS fs yet?) - the multi-OS approach and the concept of
> multitasking seem to be mutually exclusive.
Boot Shapeshifter and learn...
d.
--
Don Romero ~~~~~~~~~~~~~~~~~~~
// aka: d9...@indy.net "We don't know what we don't know."
\\// Amiga User since 1986 ~~~~~~~~~~~~~~~~~~~
Stefan Burstrom
On 30-Sep-97 16:19:52 Steffen Haeuser (mag...@birdland.es.bawue.de) wrote:
>I think it is no use, if Haage&Partner and Phase 5 say things like that to
>each other. I read your statement as to Haage&Partner wanting to overtake the
> market, i read statements from people at Haage&Partner that Phase 5 would
>want to overtake the market and would not deal good with concurrence
Phase 5 has all the way been promoting OPEN standards, like the System V
ABI, ELF etc. while H&P has been promoting their own proprietary system.
Can you explain what P5 gains from this? They are not trying to sell
software, they are trying to sell a complete system. H&P tryes to get a cut
of the software market by promoting their tools and systems as a complete
solution while P5 tries to make sure that as many developers as possible
can develop and use PowerUp by prividing Open standards and free tools.
Imagine this:
What if H&P sais 'Everyone who wants to use the extended hunk format must
licence it from us'?
Where did P5 say 'To develop for PowerUp you have to licence our software'?
That is the MAJOR difference between H&P and P5!
>situations. Such statements, i believe have no place in a business
>situation. If i compare Haage&Partner and Phase 5 statements - well, i
>believe the one of H&P was much more fair to the other firm.
>Of course you now could say "You are a Haage&Partner fan", but in fact that
>is not the case. Phase 5 and Haage&Partner are both firms for me. Phase 5
>did great hardware - Haage&Partner did great software.
I wouldn't call software that enforces people to buy H&P software 'great
software'
wd>> That means that the kernel, the (only compatible) compiler as well as
wd>> some application software - including their java development - come from
wd>> one source. They even claim their software could be a standard for
wd>> future PPC Amigas from Amiga Inc. - provided that Amiga Inc. let
wd>> themselves and their OS development be controlled by the H&P solution,
wd>> what H&P doesn't say.
wd>> Now why has Haage&Partner spent so many efforts in developing an
wd>> incompatible kernel, which is quite similar to our own system software
wd>> and offers no relevant advantages? Could it be that they want to control
>According to the programmer of WarpOS simply because the new kernel is more
>efficient. And no, he says, this is NOT based on Beta Versions and Developper
> Boards. Also the way it is included into the system would be "more OS
>legal". And if it is so similar, why do you at all have problems with it ?
If it stops me from using software I bought the CSPPC board for, it will
bug me like hell.
Also, I'd like to see an independant comparison between the PowerUp
software and WarpOS before I comment on the efficency part. I have been
working with Ralph Smidth for 3-4 months on the PowerUp software and there
has been significant speedups due to suggestions and testing from my end.
And this even without having to disassemble 1 single line of the PowerUp
kernel.
>Steffen Haeuser
Stefan Burstrom
On 30-Sep-97 10:58:00 Rudi Chiarito (chia...@cli.di.unipi.it) wrote:
>That makes no sense at all for me. A bad programmer is a bad programmer, no
>matter what tools he's using. If he's too lazy to rewrite the program to
>take full advantage of PowerUps, you can be sure he will write bad programs
>even if he's using your tools. StormC's approach makes your life easier, but
>it doesnt enforce you to adopt a lazy strategy.
>And a good programmer is always a good programmer: he won't be satisfied if
>a simple recompile gives him a program which is two times faster. He will
>always try to get the most out of his hardware.
>What you're talking about would be the developer's fault, not the
>development tool's.
If the development tools encourage peoble to write bad and slow code, more
developers will do that. Just take Visual Basic as an example. If Visual C
would have been as intuitive as Visual Basic, more programs would have been
written in C instead of Basic.
Niels Knoop
Wolf Dietrich (w...@gf.phase5.de) wrote:
> As we appreciate freedom of choice, we are happy to see these efforts
> and will support them as good as we can.
I'm really glad to hear this. Unfortunately certain moves of your
company in the past made a different impression on me. Like those:
"<Xyz> is for CyberGraphX.
You are not allowed to use it with so called CyberGraphX API
compatible clones."
It's been the same sad story with quite a number of programs
directly or indirectly controlled by Phase 5. Programs which,
after all, are usually meant as extras for customers of Phase 5
hardware - like the PowerUp boards - or even seen as tributes to
the Amiga community.
Now what about the user's freedom of choice here? Phase 5 is
actively pushing artifical (legal or technical) restrictions
which can bar a lot of users from using the RTG system of
their choice.
The mess this creates is in some ways similar to the one you
complaining of in the WarpOS issue. Picasso96 users, even
customers of Phase 5, shall be forced to switch to CyberGraphX
and reboot every time they want to try out their new PowerUp
board with some of the provided programs?
> We support any developer who wants to build a compiler (see above)
> or anything else (which is compatible).
Please try to see Picasso96 in this light. Its authors have built
something which is indeed compatible. Picasso96 does not challenge
the CyberGraphX API standard by enforcing an incompatible approach
and also does not break existing and coming software written for it.
As a potential buyer of a Blizzard 2604 board, but also a satisfied
Picasso96 user, I would appreciate it very much if Phase 5 would
stop this whole monkey business. The efforts to enforce CyberGraphX
as a monopoly were better invested into developing and enhancing it
as a general API standard and improving the implementation.
"Try our system. You'll see that it is better" is a valid argument.
"Use our system. You are not allowed to use another one" is none.
--
| Bye! /// Niels Knoop /// e-mail:
ni...@rbg.informatik.th-darmstadt.de |
--
Stefan Burstrom
On 30-Sep-97 06:14:57 George Noel (gn...@morgan.ucs.mun.ca) wrote:
>>boards are concerned, we are the system vendor, and believe that we
>>the HAL, especially to make sure that all future developments by all the
>>developers who support PowerUp remain compatible.
>But why? Competition is good.. if someone can provide a genuinely better
>software, then let them do it.. it is up to the end user to decide which
If someone provides competitive software which dissables other companies
software without providing a compatible solution, that is _very_ bad and
can never be tolerated how much better the competitive software is.
>one is the best and if they or developers are going to support it. If
>their software has compatibility issues, genuine ones, not ones forced on
>them by you, then it will be looked down upon anyways.
>>If you want to make a comparison with Microsoft, you should question the
>Yours does look like one also.
Phase5 wants to make sure that open and public solutions runs on PowerUp
that enables as many developers as possible to use it. H&P tries to enforce
people to use their solution and thier tools.
>I still don't like this monopolistic control on your companies part.
In what ways are Phase 5 monopolistic? By trying to get as many developers
as possible to get working solutions or by trying to rely on public and
well adobted standards (like elf)
Steffen Haeuser
ĸ° w...@gf.phase5.de schrieb : ĸī
Hallo !!!
wd> phase 5 supplies, as I stated, a solution which consists of hardware and
wd> software. We provide a HAL on which every developer can integrate any
wd> solution he likes, while staying compatible. As far as the PowerUp
wd> boards are concerned, we are the system vendor, and believe that we
wd> should control the system software that controls the hardware underneath
wd> the HAL, especially to make sure that all future developments by all the
wd> developers who support PowerUp remain compatible.
Well, i understand your point... but if this is your main interest, why did
you decline all cooperation with Haage&Partner ? As i was told, they offered
this once more to your firm, before they actually released WarpOS... in fact
they even did not leak ANY information about it, before they spoke to Phase 5
... i think the HAL to be used with PowerUP should be:
- efficient
- comfortable in usage
- support modern compilers
I think this should be more important than who does the software... why not
shipping WarpOS with PowerUP ? Then you would not have to call it "a foreign
solution". I am sure, Haage&Partner would not have a problem with the
permission to use it for free.
wd> If you want to make a comparison with Microsoft, you should question the
wd> strategy of Haage&Partner, who - as a software and development system
wd> vendor - want to take over full control of the hardware that we deliver.
I think it is no use, if Haage&Partner and Phase 5 say things like that to
each other. I read your statement as to Haage&Partner wanting to overtake the
market, i read statements from people at Haage&Partner that Phase 5 would want
to overtake the market and would not deal good with concurrence situations.
Such statements, i believe have no place in a business situation. If i compare
Haage&Partner and Phase 5 statements - well, i believe the one of H&P was much
more fair to the other firm.
Of course you now could say "You are a Haage&Partner fan", but in fact that is
not the case. Phase 5 and Haage&Partner are both firms for me. Phase 5 did
great hardware - Haage&Partner did great software.
wd> That means that the kernel, the (only compatible) compiler as well as
wd> some application software - including their java development - come from
wd> one source. They even claim their software could be a standard for
wd> future PPC Amigas from Amiga Inc. - provided that Amiga Inc. let
wd> themselves and their OS development be controlled by the H&P solution,
wd> what H&P doesn't say.
wd> Now why has Haage&Partner spent so many efforts in developing an
wd> incompatible kernel, which is quite similar to our own system software
wd> and offers no relevant advantages? Could it be that they want to control
According to the programmer of WarpOS simply because the new kernel is more
efficient. And no, he says, this is NOT based on Beta Versions and Developper
Boards. Also the way it is included into the system would be "more OS legal".
And if it is so similar, why do you at all have problems with it ?
Steffen Haeuser
Stefan Burstrom
On 30-Sep-97 00:19:20 Holger Metz (hm...@rz.uni-frankfurt.de) wrote:
>they didnt hack a kernel.. they made a message system like phase5
Now tell you how they figured out how the CyberStormPPC hardware works
without hacking the hardware??? Also, does anyone have a reliable speed
test? The only test I have seen so far is the Voxel engine comparison which
may be a very bad comparision since it seems to be based on older PowerUp
software and can very well be optimized for WarpOs and unoptimized for
PowerUp. (NB: I am not indicating anything, I am just asking the questions)
>library ! ..Thats the reason for the SPEED increase.. their own
>programms in elf run slower because
>they use a ppc.library from december 96.. which has this task
>switching..
>this is the reason why their prog. runs so slow ! with the newer
>ppc.library it would run
>possibly the same speed !
That is very likely since the latest ppc libraries has shown alot of speed
increas.
Can anyone write an independant speed test to figure this out? (I guess
people think that I am too closely tied to P5 to believe facts I provide)
>
>>
>> Developing the P5 way means; both free compilers and tools (gcc etc),
>> and also (possibly) the choice to use other industry-standard tools
>> that
>> can work with ELF objects. Now tell me, what way other than the
>> commercial
>> Storm-tools is there to compile his/her applications to the H&P
>> format?
>> The WarpUp docs don't give a clue, anyhow.
>>
>they give a clue !! you have includes for all assembler and c compiler
>!! THE MAIN problem isthat there are only 2 compilers which produce PPC
>.. Storm C and GCC .. ! Does this give you a clue..?
And SAS/C. Besides, what is best, base your system upon existing and
working standards (elf) or create proprietary ones (extended hunk format)
Per Jonsson
> Hello? What is going on here, are we taking the side of H&P?
> Not me, I think what they have don't is totally useless, stupid and
> a bad thing for the Amiga.
It's funny Wolf D. have wrote more on one week than most of
the majority on one month.
But teher is no response or reaction from H&P, if I got a better
product and got a lot of "shit" in a newsgroup I would scream.
I want to see some response from some at the H&P base.
> What a waste of time, energy and probably good programmers, they should
> have done something positive for the Amiga.
It's a pity. Instead of making WarpOS they could have joined P5 and
maybe optimized the CyberGL just a little bit more.
> CBM's own products would most of the time be the most compatible, using
> their own software, if you remember.
> What a shame, I like H&P for their ArtEffect effort, I want them to do
> positive things for the Amiga, not make things more difficult.
I still hope this just is a big missunderstanding. I think it would be
the best thing.
/Per Jonsson
Jeroen T. Vermeulen
In article <60ntac$jhi$8...@pania.unipi.it> chia...@cli.di.unipi.it (Rudi Chiarito) writes:
> Wolf Dietrich (w...@gf.phase5.de) wrote:
> >boards are concerned, we are the system vendor, and believe that we
>
> wrote other system software that controled the hardware, like Minix, BSD, etc.
> If a customer buys an hardware product, I guess he/she has the RIGHT to
> choose what to run on it.
Nonono. You can't let people choose for themselves. Imagine what would happen:
Some third party could then write, say, a replacement for graphics.library so it
would run with graphics cards. Of course it wouldn't be entirely compatible, so
there would then be (in Phase 5's words) a proprietary incompatible solution,
and the owner of the Amiga would be forced to change the OS to prevent this.
What did you say? Phase 5 already *depends* on such a replacement graphics
library for their business and it's called CyberGraphX? OTOH of course they
were rightfully upset when somebody else then made a third-party replacement
for CyberGraphX...
--
,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,
; Jeroen T. Vermeulen \\"How are we doing?"// Yes, we use Amigas ;
;--- j...@xs4all.nl ---\\"Same as always."//-- ... --;
;jver...@wi.leidenuniv.nl \\"That bad huh?"// Got a problem with that? ;
''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
As any salmon can tell you, smoking stunts the growth
Thomas Tavoly
In article <19971001.8...@asd01-15.xs4all.nl> j...@xs4all.nl (Jeroen T. Vermeulen) writes:
>
> Nonono. You can't let people choose for themselves.
Very true. Gateway/AI should immediately step in with some flaming swords and
announce their own proprietary custom chip ridden machine + closed OS <G>
> Imagine what would happen:
> Some third party could then write, say, a replacement for graphics.library so it
> would run with graphics cards. Of course it wouldn't be entirely compatible, so
> there would then be (in Phase 5's words) a proprietary incompatible solution,
> and the owner of the Amiga would be forced to change the OS to prevent this.
AFAIK noone ever changed the OS so it wouldn't support a replacement
graphics.library, nor did C= anything much to facilitate one. I presume you
allude to CyberGraphX here, but remember that CGFX is patching the OS in a
way that programs written to conform to C= guidelines are retargeted without
problems. Then there was EGS, which most people have probably forgotten by
now, which didn't.
If you want to draw comparisons, WarpOS seems the EGS of PPC to me, but
PowerUp is here sooner than CGFX was then (still replacing EGS).
Then there is P96, which is basically CGFX with a different twist, it aims
to replace something that is already functioning well, out of commercial
considerations. Lucky for us, both work fine with most software.
Can you say the same for WarpOS vs. the p5 HAL?
> What did you say? Phase 5 already *depends* on such a replacement graphics
> library for their business and it's called CyberGraphX? OTOH of course they
> were rightfully upset when somebody else then made a third-party replacement
> for CyberGraphX...
Sure they were, any business would be. Rightfully for P96? It's competition,
which might be viewed as good, but as far as standards go, might be not.
VT could have done almost as well with CyberGraphX, they made a business
decision not to, which is quite understandeable. They gained some time and
performance (not to mention control) and thus a better short term revenue.
For a company in the Amiga market as it was/is between 94 and now that is a
prerequisite for survival.
H&P is similarly trying to get control over the software side early on, though
maybe with a bit of a rushed system. Unfortunately it is incomparable to the
RTG situation, because the two systems are not compatible and will cause major
headaches later on.
I wonder why everyone seems to agree on an open AmigaOS, but not on equally
obvious stuff like a unified base for PPC development. Since GW/AI are still
deliberating on the colour of the cafetaria tables, the actual hardware vendor
should be a good choice to control the HAL, or at least guide cooperation,
which H&P doesn't seem to think is as profitable as going their own way.
...
_ . Thomas Tavoly
. _ // . aTm...@amiga.cistron.nl
. \X/ http://www.cistron.nl/~ttavoly
... 5.1
Mathias Ortmann
Wolf Dietrich <w...@gf.phase5.de> wrote:
>This is exactly the problem that we addressed when we questioned whether
>a second RTG-like system would make sense, or will be beneficial for the
>users, when P96 was released.
Stop making yourself look like a fool. You know _exactly_ what the primary
justification for the existence of P96 is: It has a _documented_ driver API,
allowing independent third parties to develop P96 drivers for their gfx
hardware. The phase5ish attitude of the "very engaged and enthusiastic
individual" you were mentioning earlier in your post turns CyberGraphX
into just another building block for the micromonopoly you are trying to
establish in the Amiga market.
Sheesh...
Wolf Dietrich
Niels Knoop wrote:
>
> Wolf Dietrich (w...@gf.phase5.de) wrote:
>
> > As we appreciate freedom of choice, we are happy to see these efforts
> > and will support them as good as we can.
>
> I'm really glad to hear this. Unfortunately certain moves of your
> company in the past made a different impression on me. Like those:
>
> "<Xyz> is for CyberGraphX.
> You are not allowed to use it with so called CyberGraphX API
> compatible clones."
This is not stated by us. I think it's written on the CGX homepage which
is done by a very engaged and enthusiastic individual but not directly
influenced or controlled by phase 5.
> It's been the same sad story with quite a number of programs
> directly or indirectly controlled by Phase 5. Programs which,
> after all, are usually meant as extras for customers of Phase 5
> hardware - like the PowerUp boards - or even seen as tributes to
> the Amiga community.
Well, you can see the the free CGX V3 as a tribute to the Amiga
community, as well as the software we developed for CGX V3. Take the
CyberGl library as an example: There's one optimized for CGX V3 (all for
free), and there's one which works dithered on 8-bit graphics such as
AGA.
We pay developers and programmers to build such programs, they don't
fall from heaven. I think it's understandable that we do not want to
spend more work as necessary by directly supporting different standards.
If a program works well on CGX V3 and not on Picasso96, we will not
spent efforts to make it work on P96. If a user decides for P96, a wants
to use such a program, he/she may ask the authors of P96 to provide the
same functionality.
> Now what about the user's freedom of choice here? Phase 5 is
> actively pushing artifical (legal or technical) restrictions
> which can bar a lot of users from using the RTG system of
> their choice.
They can use the system of their choice, but can they expect that the
goodies and extras that we develop and provide for CGX V3 work on P96? I
don't think so. If we extend the functionality of CGX V3 to add more
features for the benefit of the users, will we then be blamed that P96
doesn't have these features?
> The mess this creates is in some ways similar to the one you
> complaining of in the WarpOS issue. Picasso96 users, even
> customers of Phase 5, shall be forced to switch to CyberGraphX
> and reboot every time they want to try out their new PowerUp
> board with some of the provided programs?
This is exactly the problem that we addressed when we questioned whether
a second RTG-like system would make sense, or will be beneficial for the
users, when P96 was released. If you believe P96 is better and decide to
use it, can you then blame us that our advanced and latest developments
that we follow to make CGX V3 better may not work with it?
> > We support any developer who wants to build a compiler (see above)
> > or anything else (which is compatible).
>
> Please try to see Picasso96 in this light. Its authors have built
> something which is indeed compatible. Picasso96 does not challenge
> the CyberGraphX API standard by enforcing an incompatible approach
> and also does not break existing and coming software written for it.
There is a major difference. We support any developer who wants to
develop for CGX V3 - and we do it by adding more features, for example.
P96 is not a software supporting CGX V3, but a replacement for it. When
we add features to CGX V3 or supply additional functionality libraries
for it, we do not feel responsible if these features or additional
libraries can not longer be used when someone replaces CGX by P96.
> As a potential buyer of a Blizzard 2604 board, but also a satisfied
> Picasso96 user, I would appreciate it very much if Phase 5 would
> stop this whole monkey business. The efforts to enforce CyberGraphX
> as a monopoly were better invested into developing and enhancing it
> as a general API standard and improving the implementation.
That is what we did all the time.
> "Try our system. You'll see that it is better" is a valid argument.
> "Use our system. You are not allowed to use another one" is none.
As I outlined above, the second statement is not our position.
Steffen Haeuser
ĸ° w...@gf.phase5.de schrieb : ĸī
Hi!
wd> You really think this is scary? All that we want to do is to insure that
wd> on a PowerUp board, the software developed by us or by others using the
wd> different available development packages can run AT ANY TIME. We do not
wd> want to see our users getting frustrated because they have to make a
wd> plan which software they may run in parallel or not. We spend work and
wd> money in free software and services such as CyberGL native and the MPEG
wd> stuff (not to mention our comprehensive PowerUp system software), we
wd> provide a free development system, and yes, we also gave out certain
wd> free developer hardware. Isn't it really scary that H&P follows a path
wd> that disables all the results that come out of these efforts? I believe
wd> so. We are not against compatible alternatives, and we supported H&P and
wd> their efforts all the time while they always promised us their stuff
wd> would be completely compatible with ours - until, surprise surprise,
wd> last week.
Well, i see your point. Well, about the speed discussion, i think, best would
be, if a benchmark (or a demo like that Voxelspace thing from H&P) would be
directly ported to the "other" system to have a direct comparision. All other
discussion is useless.
Well, BTW, after Phase 5 said that ppc.library has PPC Native Multitasking,
now could you REASONABLY explain to me what this 68k task "PPC Schedule
Handler Process" is doing ? (Well, and don't tell me, i would not be allowed
having a look at a process-list program...)
I would really like the idea of a program running on both WarpOS and
ppc.library. In fact i have some programs that could be used for this, myself
(of course the sources had to be released, so that it is visible for
everybody, that the "favored" solution of the programmer was not better
optimized...). It should be done in a way, that the ppc.library part really
uses ppc.library, not WarpOS V7 (even, as far as i know it, this would not
result in a speed difference... you said it would, i would be open to an
explanation in technical terms).
wd> ??? Do you really think that's a solution? I can already see the flame
wd> and hate postings from users who would have to reset their system when
wd> they want to use a PowerUp compliant program after they have started a
wd> WarpUp-based program before.
Well, yeah, this is quite a problem. Hmmm, Haage&Partner claimed you declined
all offers of cooperation. What is this about it ?
wd> You may let users of the PowerUp boards speak for themselves, if they
wd> want a software that disables many of the applications and the utility
wd> and library software that they got with their boards or buy from other
wd> vendors.
Well, i have to admit, currently i know of more projects using the H&P
Solution (or even "patches" for ppc.library to make it more efficient) than
native ppc.library using projects... at this place i want to say something
important. I find it a VERY bad thing for the USERS if you try WarpOS V8+
not to run any longer. This would not prevent WarpOS from running (the
wrap-around on ppc.library will still be running), but it will be a BAD thing
for the users.
>> >further. Additonally, soon an update for SAS/C on the Amiga can be
>> >expected, which supports PowerUp and the ELF format.
>>
>> Hey.. this is great news!
wd> As we appreciate freedom of choice, we are happy to see these efforts
wd> and will support them as good as we can.
Well, what about a compiler that supports both formats ? Now THAT would be an
idea :) (At least, until this stuff settles...)
wd> We support any developer who wants to build a compiler (see above) or
wd> anything else (which is compatible). Looks like there will soon be two
wd> free products on the market. The H&P compiler costs a lot of money, and
wd> people may be forced to buy it to develop under WarpUp. We do not want
As H&P stated it, they were unhappy with the way things run. They would have
been more than happy with some sort of "compromiss" which would ensure
compatibility, they said. They said, all sorts of compromiss suggestions where
declined by Phase 5. I would appreciate an explanation.
wd> As a matter of fact, I believe this whole mess started in the early
wd> times when we were looking for a powerful PPC compiler to support
wd> PowerUp. We decided to use the ELF binary format because it opens up
wd> many options, such as using powerful compilers available on other
wd> systems or free compiler such as the GNU. H&P *always* wanted to push
wd> their "extended-Hunk-Format" and make the Storm C the "official" and
wd> only PowerUp development tool - but of course not fro free...
As i understood it, they wanted other compilers using it too... they only took
it for their own, as you declined their cooperation offer, they said (of
course i cannot judge, if they said the truth here... but to hear both sides
as to this argument, would be... interesting).
wd> ??? H&P told us that it is not compatible with our system software
wd> solution.
It is the way, that V8 is not compatible with your software. And V8 is it,
which gives speed enhancements over ppc.library... (i know this words could be
considered harmful... sorry...)
wd> Do we hinder them to speak? We just bring our arguments, as H&P does.
wd> And we protect those developers who develop on top of our software
wd> solution by making sure their software runs on our boards. Just as a
wd> side note, imagine how much support trouble can be caused if the H&P
wd> software disables ours for the software vendors as well as us.
They never tried anything to disable it. In fact they tried to cooperate and
even did not tell the public about it until very last. At least that is what
they tell. They told me, before they finally released it, they contacted you
AGAIN for the wish of cooperation and compromis, but that they would have been
declined COMPLETELY again. Of course this is what they say.
wd> We can make sure that compatible software will work on our boards, and
wd> H&P may make sure that their software is compatible or not.
Could it not be, that this incompatibility is a direct consequence of the
decline of cooperation ? After what Haage&Partner says (what they say, not me
!!!) the two solutions could have been joined in the way of developpement, in
a way, that then all two compilers (and if SAS comes out, all three) would
support the ONLY existing solution than that would have been partially done by
both firms.
wd> It is a conceptual problem and covers issues such as programming style.
wd> Please see the attachment below, which is a copy of a posting to this
wd> newsgroup from yesterday.
Well, i will read it.
wd> ?? Which falling out? There's a similar situation: ProDAD always wanted
wd> to convince us that their product is so great and we should adopt it as
wd> a standard, as H&P did in the past nine months. Well, we couldn't see
wd> the advantages of their concept (and in my personal opinion, their pOS
wd> is - even on 68k - still far from being a complete OS rather than a
wd> replacement for some GUI and middleware parts of the Amiga OS)
Well, at least on p-OS i share your doubts...
wd> Then you should ask them to make their JAVA compatible with our PowerUp
wd> system software, so that their JAVA and e.g. a VRML plugin for a Web
wd> Browser which uses our PPC-native CyberGL library would work together.
wd> What would YOU say if these two products would only work exclusively?
wd> Whom would you blame?
Both firms, probably.
wd> What do you mean? We have developed a very complex hardware concept, and
wd> the first product of this new line is out now. Other products will
wd> follow, which may offer different options. To make sure that software
wd> will be compatible, we provide a HAL. From that level on, it is an open
wd> system. Software such as WarpUp that hacks the hardware may not be
wd> compatible with other baord version when there are new developments and
wd> hardware or revision changes. And the hardware design of our own
wd> products WILL remain under our control.
Then you want to punish the users for the squarrel that happened as
Haage&Partner and Phase 5 did not find to cooperation ? That does not make
your point more believable for me.
wd> As a matter of fact, someone must control a system (well, not really
wd> someone but somesoftware ;-)) Why should it be the software that has
wd> been hacked by a third party, who has reengineered our software, does
wd> not know about the full functionality and options of the hardware, and
wd> who only claims his software would be better than ours in a massive
wd> marketing campaign, which very directly indicates that our software
wd> would not be sufficient?
I heard the same statement from independent sources.
wd> Yes, and the demo program running under the PowerUp System Software has
wd> also been provided by them. Oh, by the way, was it the "Set Pixel"-Demo
wd> which they have shown you? If so, this demo is as useful as testing a
wd> Porsche by driving backwards.
No, it probably was the Voxelspace demo, he was talking about. (Do you REALLY
had a look at WarpOS yourselves? Well, i know that you are the main guy at
Phase 5, but before telling things about it you should have looked at it at
least...)
wd> I am not a software developer, and can just explain it with the basic
wd> understanding that I have, so please excuse if this is not going too
wd> much into the details. We at phase 5 digital products do recommend that
Well, you have software developpers how COULD do this at hand, don't you ?
wd> software developers who want to support PowerUp rewrite their programs
wd> in a way that they build software modules in form of tasks which run on
wd> PPC and 68k. In this way, the software is well structured, open for
This sounds like the "mirror process" solution which is the way WarpOS is
implemented...
Well, i will reply to the rest of the mail (regarding the more technical
discussion) in a different mail to make this one not too long.
Steffen Haeuser
Steffen Haeuser
ĸ° gene_h...@iolinc.net schrieb : ĸī
Hi!
ge> However, lets put this in the proper perspective. Will your new car
ge> work equally well if instead of 87 octane, you put some diesel fuel
ge> of a reasonable cetane number it there instead? Note that I said
ge> 'work equally well' there. Yes, it will run, but not quietly, nor
ge> with much power, prompting you to either run it on through, driving
ge> very gently, and diluting it with the best gasolene sold everytime
ge> theres room for another gallon in the tank, or have the tank drained.
ge> Been there, done that, tanker filled wrong tank with super-diesel!
ge> If Wolf wants it to run his grade of fuel, and you don't agree about
ge> that attitude, then either get over it, or go buy somebody elses.
ge> What, nobody else is making this model? Then get over it.
Well, it is more that we get the new car and the firm producing the new car
wants us to use Diesel fuel for it (even if there is 87 octane available), and
then they start to make all tank stations trouble who sell 87 octane.
At least currently it looks like that to me. As soon as i get my PPC Board i
will do some really comparision test, this is sure.
Steffen Haeuser
Martin Skowronski
>Mathias Ortmann (ort...@informatik.tu-muenchen.de) wrote:
>: Wolf Dietrich <w...@gf.phase5.de> wrote:
>: This is ironic. Even Microsoft pales beside this kind of business practice.
James McArthur made the following observation:
>Seems kind of silly thought that H&P would stab P5 in the back since P5 are
>currently the only Amiga PowerPC vendors.. makes me wonder what we _arent_
>being told.. (cue conspiracy theory.. :>)
Personally, I think we are being "told" too much ;-) You'll never see any
Microsoft/IBM/Apple/Gateway etc. top executives engage in this type of inflammatory
bickering.
If Wolf Dietrich feels that WarpUP is either incompatible or squanders the
potential of the PowerUP cards; then he should simply make an announcement to that
effect. phase5 has not benefited in the least by Mr Dietrich's recent postings
concerning WarpUP. In fact, his statement's have only served to undermine the
stability of the Amiga market as a whole.
H&P and phase5 *need* to resolve their differences in private. Heck, if Bill
Gates and Steve Jobs can waltz on stage hand in hand, surely Wolf Dietrich
can come up with a more *benign* way of announcing his disapproval of WarpUP.
Martin Skowronski
ProActive Images
www.proactive.net
PGP public key on request
Wolf Dietrich
Steffen Haeuser wrote:
> Well, BTW, after Phase 5 said that ppc.library has PPC Native Multitasking,
> now could you REASONABLY explain to me what this 68k task "PPC Schedule
> Handler Process" is doing ? (Well, and don't tell me, i would not be allowed
> having a look at a process-list program...)
kernel, which of course does the scheduling and runs in native PowerPC
code (what else?).
> Well, yeah, this is quite a problem. Hmmm, Haage&Partner claimed you declined
> all offers of cooperation. What is this about it ?
Read my comment concnering cooperation below, please.
> I find it a VERY bad thing for the USERS if you try WarpOS V8+
> not to run any longer. This would not prevent WarpOS from running (the
> wrap-around on ppc.library will still be running), but it will be a BAD thing
> for the users.
Once again, we ONLY want to make sure that customers of our products can
use software based on the PowerUp System Software and the PPC library at
any time. If this means that WarpOS V8+ may not run, it is a consequence
of the design that H&P have been choosen.
> Well, what about a compiler that supports both formats ? Now THAT would be an
> idea :) (At least, until this stuff settles...)
And what about very powerful compilers for PPC that run on other
systems, generate ELF format binaries, and can provide the possibility
to compile highly optimized code which may be far more efficient than
the solutions available on the Amiga now? We at least want to have this
option to be able to compile our software for highest performance and
optimized code.
> As H&P stated it, they were unhappy with the way things run. They would have
> been more than happy with some sort of "compromiss" which would ensure
> compatibility, they said. They said, all sorts of compromiss suggestions where
> declined by Phase 5. I would appreciate an explanation.
Read my comment concerning cooperation below, please.
> wd> As a matter of fact, I believe this whole mess started in the early
> wd> times when we were looking for a powerful PPC compiler to support
> wd> PowerUp. We decided to use the ELF binary format because it opens up
> wd> many options, such as using powerful compilers available on other
> wd> systems or free compiler such as the GNU. H&P *always* wanted to push
> wd> their "extended-Hunk-Format" and make the Storm C the "official" and
> wd> only PowerUp development tool - but of course not fro free...
>
> As i understood it, they wanted other compilers using it too... they only took
> it for their own, as you declined their cooperation offer, they said (of
> course i cannot judge, if they said the truth here... but to hear both sides
> as to this argument, would be... interesting).
The GNU C for PowerPC supports the ELF format, as well as other
compilers do. We don't write compilers, H&P does. We simply take a
compiler which is available and appropriate, while H&P are able to
modify theirs to support other standards. Why should other compiler
makers follow the H&P standard? And would you expect software vendors
that built VERY professional compilers with prooven performance,
functionality and optimization which run on other systems that they
built a version to support the H&P standard? This, BTW, is also valid
for other kinds of software.
So, do we really need to have a proprietary binary format that EXISTS
ONLY on PowerUp-Amigas, respectively? I don't think so.
> It is the way, that V8 is not compatible with your software. And V8 is it,
> which gives speed enhancements over ppc.library... (i know this words could be
> considered harmful... sorry...)
As you said yourself, independent tests would have to proove which
software and which design concept/recommendation is faster. Until this
has been done, I think it is not correct to make statements as you do,
which are written in a way as it would be a fact that WarpUp is faster
or has any other advantages.
> They never tried anything to disable it. In fact they tried to cooperate and
> even did not tell the public about it until very last.
??? They built a software that disables ours, and they worked on it
secretly for nearly a year while appreciating our developer support.
This is not what I assume to be cooperative.
Read my comment concerning cooperation below, please.
> At least that is what
> they tell. They told me, before they finally released it, they contacted you
> AGAIN for the wish of cooperation and compromis, but that they would have been
> declined COMPLETELY again. Of course this is what they say.
We told them that we completely disregard a software that disables the
PowerUp System Software because of the reasons that we publicly
explained.
Read my comment concerning cooperation below, please.
> wd> We can make sure that compatible software will work on our boards, and
> wd> H&P may make sure that their software is compatible or not.
>
> Could it not be, that this incompatibility is a direct consequence of the
> decline of cooperation ? After what Haage&Partner says (what they say, not me
> !!!) the two solutions could have been joined in the way of developpement, in
> a way, that then all two compilers (and if SAS comes out, all three) would
> support the ONLY existing solution than that would have been partially done by
> both firms.
Read my comment concerning cooperation below, please.
> wd> What do you mean? We have developed a very complex hardware concept, and
> wd> the first product of this new line is out now. Other products will
> wd> follow, which may offer different options. To make sure that software
> wd> will be compatible, we provide a HAL. From that level on, it is an open
> wd> system. Software such as WarpUp that hacks the hardware may not be
> wd> compatible with other baord version when there are new developments and
> wd> hardware or revision changes. And the hardware design of our own
> wd> products WILL remain under our control.
>
> Then you want to punish the users for the squarrel that happened as
> Haage&Partner and Phase 5 did not find to cooperation ? That does not make
> your point more believable for me.
We punish the users? That's weird. We simply provide a complete
solution, and we provide a HAL. Below the HAL, there is the hardware,
which is of course our proprietary development and may change according
to actual needs of upcoming product developments (such as other PowerUp
boards). This will not affect the software developers who use the HAL,
and also not the users who work with this PowerUp compliant software.
> No, it probably was the Voxelspace demo, he was talking about. (Do you REALLY
> had a look at WarpOS yourselves? Well, i know that you are the main guy at
> Phase 5, but before telling things about it you should have looked at it at
> least...)
I had a look on it, and I don't find the Voxelspace demo so impressing,
and also not it's speed increase over the 060 version. I also don't find
their Mandelbrot impressing, and I can't see that it is faster than our
Mandelbrot demo at all. Have you seen our real-time CyberGL-based
Lightwave object viewer? Have you seen our MPEG Audio/Video player
operating, even in it's first unoptimized release? I must say, frankly,
that I believe that I have seen more of both solutions right now than
you did.
> wd> software developers who want to support PowerUp rewrite their programs
> wd> in a way that they build software modules in form of tasks which run on
> wd> PPC and 68k. In this way, the software is well structured, open for
>
> This sounds like the "mirror process" solution which is the way WarpOS is
> implemented...
Didn't I say that the H&P WarpUp kernel does not provide anything new? I
read their description, and their task and message functions, for
example, are a similar copy of the functionalities that our PowerUp
System Software provides - even the names are equal or nearly equal.
A special final comment on cooperation: Your post may give a strong
impression to readers that phase 5 digital products is not cooperative.
It states at various points that we had declined cooperation. This is a
statement that expresses the position of Haage&Partner, and it must be
seen as a subjective statement. Facts, however, proove that we have
given them all necessary developer support to become a PowerUp
developer. H&P may be of the opinion that we have been uncooperative
because they could not convince us to use their binary format and make
their compiler the development standard. We believe this is a very
personal point of view, and that by repeating this statement they try to
shift the blame for the confusion that their confrontational course may
caused towards us.
Please excuse if I may drop out of the public discussion at this time,
as business duties will keep me more busy - we want to continue bringing
powerful products to the Amiga market. We will probably address this as
well as future issues with more information about our products and the
concepts behind on our web site.
Steffen Haeuser
ĸ° chia...@cli.di.unipi.it schrieb : ĸī
Hi!
ch> Do you consider that a major problem? Nothing is going to prevent users from
ch> "upgrading" their Flash-ROMs with the SW of their choice. Be it from Phase5,
ch> Haage&Partner or anyone else. A plain and simple solution :)
:) Well, but i find it a VERY sad thing, that people HAVE TO DO THIS. They
should be able to choose THEMSELVES which software to use, without having such
things to do...
ch> When the user has bought your hardware, he/she can do whatever he/she likes
ch> with it. He can even play cricket with the board: you can't do anything to
ch> prevent that from happening.
:))))
Steffen Haeuser
Claude Mueller
Hallo w...@gf.phase5.de (Wolf Dietrich)
Du hast am 30.09.1997 folgendes unter "Re: The Wolf is released! (was Re: WarpOS released !!!)" von Dir gegeben:
> the advantages of their concept (and in my personal opinion, their pOS
> is - even on 68k - still far from being a complete OS rather than a
> replacement for some GUI and middleware parts of the Amiga OS)
But let's say their pOS will be a complete OS someday. Why shouldn't it
run on PowerUP ? That can only mean sellign more boards to you. You
could have your software take care of the people that use AmigaOS,
and proDad could write the drivers for their pOS for PowerUP.
As for the Haage&Partner thing, I agree with you. But pOS is/will be
a seperate OS, not to be compared with your approach.
Of course when you look at the future, pOS will probably compete with
an ABox-OS or Gateway-AmigaOS, I guess that's why you don't like pOS.
Damn, the Amiga-market is small enough, we can't afford to have
three different OSs...
Grüsse, Claude
_________________________________________________________________
//|Amiga User Group Switzerland http://www.limmat.ch/augs |
// |A2000 WILDFIRE 060/33MB QAtlas VLAB ISDN PPak T563T A10 SD64 CD32|
\X/ |http://www.eye.ch/~claudem cla...@eye.ch|
-----------------------------------------------------------------
>>> Every little thing's gonna be alright <<<
Steffen Haeuser
ĸ° mo...@bulldozer.rydnet.lysator.liu.se schrieb : ĸī
Hi!
mo> No, no, no. Running WarpOS means that P5 software, or software created
mo> with P5's _free_ devkit won't run anymore. Such a solution is a big step
mo> backwards, back to the good old days when you had to reset the computer
mo> between running programs.
Only, if you use the WarpOS V8+. WarpOS V7 (a wrap-around on ppc.library) will
always be available, and it runs parallel with the Phase 5 things.
Steffen Haeuser
Steffen Haeuser
ÿ° mil...@geocities.com schrieb : ÿ´
Hi!
mi> They are fragmenting the Amiga community even more with their obsolete
mi> OS. You won't gain anything from using their system, they too have the
mi> delay when switching processors (read the docs).
But then, this delay is even worse on the system of Phase 5 (only that they
try to discuss it away...)
mi> Besides that, the tone of their guide explaining how to make games
mi> not only sounds patronizing, but it also tells you that you have to
mi> use their system for making games or else it won't work well.
Well, the problem is, when you do games with the Phase 5 systems you will get
longer delays.
mi> Then they are telling you how fantastic Steffan Haussers system is
At least you could spell my name correctly :)
mi> for making PPC games, has it already been tested with PPC or are they
mi> just trying to get him on their side as a good advocate ( I don't mean
mi> this negatively about Steffan, I like the guy, but stay on P5's side).
There was no need for this. I was already on their side because of technical
reasons, before. But of course i appreciate it.
mi> Why in the world would they make a system that is equal to P5's one?
It is *not* equal.
Steffen Haeuser
Steven Matty
I've been reading this thread with interest, and thought I'd add
my bit...
Good Things(tm) about WarpOS
----------------------------
1) It is (allegedly) faster
Bad Things(tm) about WarpOS
---------------------------
1) Programmers wishing to write for it *MUST* use H&P's commercial
C compiler
2) It is completely incompatible with the DE-FACTO PowerUP kernel
3) It won't be possible to write code with GCC or SAS/C and run
these programs nuder WarpOS (afaik, WarpOS will not run programs
compiled to use Phase5's library in ELF format)
4) Having to reboot to run a PowerUP program or a WarpUP program
is just plain awful
5) It's a kludge - no matter how you present it, it is based on
reverse engineering of the PowerUP hardware - and is more
likely to break on future (or current?) cards.
The way I see it - it's a thinly disguised attempt to force
Amiga developers to use the StormC compiler as "The Standard"
which IMHO is a VERY BAD THING INDEED(tm)
I hope the majority of developers see sense and refuse to
fall for it.
Regards,
Steve.
.---------------------------------------------------------------------------.
| Red When Excited Ltd Telephone +44 1302 890000 to order your copy |
| TEL: +44 1705 641462 of the Blitz Support Suite NOW! |
+---------------------------------------------------------------------------+
| A1200T/16Mb/40Mhz040/6GbHD/1.76MbFD/33.6kModem/8xCD/MultiSync/HyperCOM3 |
| EMAIL: Big.Wil...@ldngedge.demon.co.uk |
| WWW : http://www.ldngedge.demon.co.uk |
`-- ---- -- ----------------------------------------------------------------'
Randy Vice
On Wed 1-Oct-1997 10:41a, mar...@netrox.net wrote:
<snip>
m> If Wolf Dietrich feels that WarpUP is either incompatible or squanders the
m> potential of the PowerUP cards; then he should simply make an announcement
m> to that
m> effect. phase5 has not benefited in the least by Mr Dietrich's recent
m> postings
m> concerning WarpUP. In fact, his statement's have only served to undermine
m> the
m> stability of the Amiga market as a whole.
Not even that, let the consumer find out which one is faster. Unless of
course, P5 comes out on the bottom...
<snip>
Mathias Ortmann
mag...@birdland.es.bawue.de (Steffen Haeuser) wrote:
>Well, Sam Jordan said, he is already working on, that they will run again with
>the changed Boards... (appearently Phase 5 have changed some things, again).
>He said he will soon be done with this.
At least you can't claim that the Amiga's agony is boring.
Sit back and enjoy the show!
Steffen Haeuser
Hi!
Now let's seriously discuss those Wolf Dietrich Statements... don't be exited
and call this a flame. It is not. I am interested in a serious discussion,
only. If you want this too, discuss technical details or let a programmer of
your firm do this. Don't simply call me "a H&P promoter". I am interested in
both software solutions. But i am not interested in "vapor features" and
"vapor speed". I am only interested in REAL FEATURES and REAL SPEED. I am open
to any critics to this article. Reply private or on the net, anyways.
>future developments and object oriented. A simple program for example
>could consist of a 68k task which handles the GUI and OS calls, and
>calls PPC tasks for executing operations which shall be accelerated by
>using the PPC. As we provide comprehensive functionality and a very fast
>new message system in our PowerUp System Software, this solution is very
>efficient, and it allows full parallel operation of multiple CPUs; in
>the example above, the 68k task could continue with execution of
>commands while one or more PPC tasks are busy with other stuff, and
>wouldn't need to wait until the PPC task(s) returns (unless, of course,
>the 68k task would need the result of the PPC task(s) to continue it's
>work).
The problem is the "Cache Problem". Your 68k tasks would NEVER be completely
independent of the PPC Tasks. In fact, they all have to share information. Let
us think about an example (i was thinking of a game/demo). I have to note, i
did this thoughts/calculations open minded. But (if i do not have made a
logical mistake) sadly there is *no* possibility to prove the Phase 5
Statements. I would also like that 5 microseconds for a message would be true,
not 0.5 milliseconds (BTW, as to Phase 5s software i would be happy, if it at
least did send a message in those 0.5 milliseconds... but appearently, it
currently behaves worse). I am open-minded to ANY corrections in these
thoughts, even making them complete nonsense. If you (or anybody else)
have/has corrections, please post them to this newsgroup.
68k task 1 : Gets Keyboard/Mouse Input and sends it to the PPC (using the
messaging system) (Task 3)
68k task 2 : Gets the current state of the game from the PPC Task 13and
sends needed data through the serial interface for Multiplayer.
68k task 3 : Handles the sound part, is activated by PPC task 3.
PPC task 1 : Gets the Data and, based on it, renders the next frame
PPC task 2 : Gets the rendered data and does the display (writing directly
to Video RAM, so that no 68k functions are needed)
PPC task 3 : The main task
I guess this is what you had in mind ?
Well, now the problem. For example 68k task 1 is NOT independent of PPC task
3. When they exchange data using the messaging system, a Cache Flush of the
PPC is needed (else you will end in datalosses, i won't get into detail
here... the docs of WarpOS and ppc.library explain this both). While this
Cacheflush is done, all PPC Tasks come to a halt. Now use stated before your
new messaging system would take "at worst" 5 microseconds for a message
exchange. Let us calculate some numbers:
the PPC has 32 KB Data Cache. 200000 times 5 microseconds would be a second.
So we need to flush 32768*200000 Bytes per second. That would be:
6.553.600.000 Bytes. This would mean, to make the time you announced you would
need a memory interface that can do 6.5 GIGABYTE / second. I don't think
PowerUP can do this. In fact i don't think ANY computer on this planet can do
this.
If there is a mistake in my calculations, please correct me. But i do not
think so.
Now let us calculate the same with the statement of what Sam Jordan told about
the WarpOS V8+ (he said on the PPC604e it would need 0.4-0.5 ms for a message).
0.5 ms (worst case) would be 32768*2000 Bytes per second. This would mean we
ended up with 65 MB/second, what sounds to me like a much more reasonable
value.
So what have we learned this far:
1. Every message to be sent will need MORE than 5 microseconds.
2. The different task are not completely independent (everytime they exchange
data, they will "meet").
It was argued, that forwarding 68k Interrupts to PPC Interrupts would help (i
believed this myself, at first), but: An interrupt only tells you that
"something happened" (for example: "A key was pressed"). It does not tell you
WHICH key. You have to get this yourselves. Okay, in cases like Keyboard
presses you now could get the information with PPC Native code from the
memory. BUT this method can only be used, if hardware interfaces (like
0xbfec01, where the last pressed key is found) are accessed. It CANNOT be used
to access data written there by 68k tasks (then again, the Cache problem
arises).
So we come to the following end:
- Interrupts could be used to enhance the speed of stuff like the
Keyboard/Mouse Input, but ONLY THEN, if Phase 5 release the hardware
data of how to access them OR if they implement a Patch for some
Keyboard/Mouse function by the OS, running PPC Native (Keyboard Handler....)
It should be also noted, that Keyboard/Mouse checks only make a VERY
small time in the overall performance. Not important, you might say.
The actual rendering takes much more time than 0.5 milliseconds.
- Interrupts CANNOT be used to enhance 68k/PPC Messaging.
>By using such a structured and object oriented programming approach, the
>software is also much more easily updatable. If e.g. GUI and OS calls
>would change in future revisions of an OS, the developer would only need
>to replace the module that handles the GUI and the OS calls, and
>continue to use the other modules.
Well, the following stuff is not that objective, like above, it is my personal
opinion:
Personally i have to say i prefer not being forced to deal with different
tasks (if i want this, i still can do it). Also, all 68k calls in PPC code
have to be done inside a RunOS() (or what it was called) function of
ppc.library, while the compiler deals with this automatically with the Storm
Solution. Also, ports are much more easy with WarpOS.
This is NOT talking about "straight ports". It is only, that "what is
possible" already is dealt compiler-internal with WarpOS. Of course extra
things can also be called.
The "object oriented" argument sounds like a half-argument to protect one's
solution to me.
>H&P have been promoting their Storm C with a concept that you simply
>recompile the existing C code which has been written for 68k Amiga and
>tell the compiler to generate PPC code. This may be very comfortable for
>the developer, but is completely unefficient. Not only that in the worst
Currently it is more efficient than ppc.library, prove your statements in own
developpements before you state them !!! Also, it is, like outlined above, no
streight recompile. Internally WarpUP works with tasks that talk to each other
(so-called mirror tasks). But (as to what i outlined above) only one of them
can be active at a time, the 68k one, or the PPC one.
>case you have a random mix of 68k and PPC code, but also the two CPUs
<will not run in parallel as it is one continuous program flow which
>jumps between PPC and 68k. If a linear code sequence is executed, the
>68k and the PPC always work one after the other, in the worst case with
>lots of switches and a useless overhead.
Your statements about "parallel" still have to be proven. The calculations i
did above seem to indicate that it is not possible.
Steffen Haeuser
Wolf Dietrich
Claude Mueller wrote:
>
> Hallo w...@gf.phase5.de (Wolf Dietrich)
>
> But let's say their pOS will be a complete OS someday. Why shouldn't it
> run on PowerUP ? That can only mean sellign more boards to you. You
> could have your software take care of the people that use AmigaOS,
> and proDad could write the drivers for their pOS for PowerUP.
I have no problems if they port pOS to PowerUp or later the A\BOX. All
we said was that it will not be our choice as a primary operating system
or solution.
> Of course when you look at the future, pOS will probably compete with
> an ABox-OS or Gateway-AmigaOS, I guess that's why you don't like pOS.
As I said, if Amiga Inc. releases a PPC version of AmigaOS, we would
surely apply for for a license if they'd offer such.
> Grüsse, Claude
Gruß zurück, Wolf (dies dämlichen Umlaute) <-German joke;-)
Niels Knoop
Wolf Dietrich (w...@gf.phase5.de) wrote:
> > "<Xyz> is for CyberGraphX.
> > You are not allowed to use it with so called CyberGraphX API
> > compatible clones."
>
> This is not stated by us. I think it's written on the CGX homepage
> which is done by a very engaged and enthusiastic individual but not
> directly influenced or controlled by phase 5.
O.K. So using Benoit and other Phase 5 programs with Picasso96 is
not illegal and there are no artificial protections to restrict them
from running on non-CyberGraphX systems, right? I'm happy to hear
this.
> Well, you can see the the free CGX V3 as a tribute to the Amiga
> community,
Yes, I do. CyberGraphX V2 already was one, even though it was not
free, because it created sort of the RTG standard which Commodore
failed to deliver.
> as well as the software we developed for CGX V3.
I agree with that as long as that software is developed for, let's
say, "the open CyberGraphX API standard V3" rather than "for our
proprietary CGX V3 RTG implementation, using undocumented features
and details".
> We pay developers and programmers to build such programs, they don't
> fall from heaven. I think it's understandable that we do not want to
> spend more work as necessary by directly supporting different
> standards. If a program works well on CGX V3 and not on Picasso96,
> we will not spent efforts to make it work on P96. If a user decides
> for P96, a wants to use such a program, he/she may ask the authors
> of P96 to provide the same functionality.
This sounds perfectly reasonable. I'm glad to hear this from you and
hope that the future moves of your company will be consistent with
this favourable attitude.
> They can use the system of their choice, but can they expect that
> the goodies and extras that we develop and provide for CGX V3 work
> on P96? I don't think so. If we extend the functionality of CGX V3
> to add more features for the benefit of the users, will we then be
> blamed that P96 doesn't have these features?
It depends. Of course, neither Phase 5 nor the CyberGraphX authors
can be blamed if Picasso96 lacks features or doesn't implement them
in a compatible way. However, they do indeed deserve the blame if
the only reason for this is that the specification is kept a secret,
or - even worse - if the programs in question do not follow the
documented API themselves.
> > The efforts to enforce CyberGraphX as a monopoly were better
> > invested into developing and enhancing it as a general API
> > standard and improving the implementation.
>
> That is what we did all the time.
Well, IMHO creating and maintaining a standard requires that its
specification and especially its API are open to the public. What
about CGX V3? The only freely available developer documentation
on Aminet and the CyberGraphX homepage still dates from January '96
and thus doesn't give any details about V3.
Steffen Haeuser
ÿ° gn...@morgan.ucs.mun.ca schrieb : ÿ´
Hi!
>>Of course, as the coder of ppc.library outlined, it is not completely the
>>fault of the software... 68k/PPC switch is a hardware problem, also. But
gn> Funny.. you wouldn't say from one of Wolf Dietrich's recent posts.. he
gn> claims the boards are fully multiprocessing and parallel.
About this, have a look at my most recent post ("ppc.library/WarpOS
discussion", or something like this i called it). You are invited to look for
logical bugs in the argumentation.
Steffen Haeuser
Steffen Haeuser
ĸ° name...@yme.himolde.no schrieb : ĸī
Hi!
>>: <Laire> sotto: warpup mag vielleicht auf den bis jetzt ausgelieferten boards
>>: laufen...das wird sich ab uebernaechste woche aendern
>>
>>: (Translation: "WarpUP may be compatible with the current boards, but this
>>: will change in a week or two.")
>>
>>They make the hardware, they should be able to say what runs and what doesnt.
na> Are you nuts? What if they go bankrupt? What if I want to run some
na> other OS? What if I install some OTHER hardware? Seems to me that they
na> have taken all teh bad sides of the Amiga and magnified it.
Well, Sam Jordan said, he is already working on, that they will run again with
the changed Boards... (appearently Phase 5 have changed some things, again).
He said he will soon be done with this.
Steffen Haeuser
Steffen Haeuser
ĸ° f94...@efd.lth.se schrieb : ĸī
Hi!
f9> As long as software stay compatible anyone can do what they want. H&P's
f9> solution is neither compatible nor very efficient.
Correction: It is not compatible and very efficient.
f9> The difference is that the PowerUp software is based on industri standard,
f9> has remained compatible with the initial releases given out to developers.
f9> H&P's solution disables the PowerUp software and makes 100% compliant
f9> PowerUp software break.
If we would all use Industry standards, we would not own an Amiga, but a Win95
or Linux system. :)
Steffen Haeuser
DATE: 1.Okt.1997 / TIME: 14:35:45
Steffen Haeuser
ĸ° chia...@cli.di.unipi.it schrieb : ĸī
Hi!
>>the developer, but is completely unefficient. Not only that in the worst
>>case you have a random mix of 68k and PPC code, but also the two CPUs
ch> AFAIK with StormC the programmer has control over which parts use PPC code
ch> and which don't...
Exactly this is the case (i am a tester of StormC 3.0 PowerUP, just to note
this).
Steffen Haeuser
Ben Hutchings
In article <6000040623...@BIRDLAND.es.bawue.de>,
Steffen Haeuser <mag...@birdland.es.bawue.de> wrote:
<snip>
>Well, now the problem. For example 68k task 1 is NOT independent of PPC task
>3. When they exchange data using the messaging system, a Cache Flush of the
>PPC is needed (else you will end in datalosses, i won't get into detail
>here... the docs of WarpOS and ppc.library explain this both). While this
>Cacheflush is done, all PPC Tasks come to a halt. Now use stated before your
>new messaging system would take "at worst" 5 microseconds for a message
>exchange. Let us calculate some numbers:
>
>the PPC has 32 KB Data Cache. 200000 times 5 microseconds would be a second.
>So we need to flush 32768*200000 Bytes per second. That would be:
>6.553.600.000 Bytes. This would mean, to make the time you announced you would
>need a memory interface that can do 6.5 GIGABYTE / second. I don't think
>PowerUP can do this. In fact i don't think ANY computer on this planet can do
>this.
This calculation is wrong because it assumes the data cache is full. If
you do lots of messaging in this inefficient way then the data cache is
unlikely to fill up much. But this seems stupid anyway.
Why can't messages be passed in shared memory marked as non-cacheable on
both 680x0 and PPC sides?
--
Ben Hutchings, compsci&mathmo | ICOAmiga http://www.lapcopaintball.com/icoa/
email/finger m95...@ecs.ox.ac.uk | homepage http://users.ox.ac.uk/~worc0223/
Man invented language to satisfy his deep need to complain. - Lily Tomlin
Stefan Burstrom
On 30-Sep-97 21:21:21 Goswin Brederlow (goswin.b...@student.uni-tuebingen.de) wrote:
>Wolf Dietrich <w...@gf.phase5.de> writes:
>> George Noel wrote:
>> > > As it does not use the HAL that is provided by
>> > >phase 5 digital products, but hacks the hardware by using an incomplete
>> > >knowledge of the hardware functionality - which has been aquired by
>> > >reengineering the phase 5 digital products software - it is most likely
>> > >that it will not work with future revision changes or different product
>> > >releases of the PowerUp series.
>It doesn't use the HAL? I heard something very different. And if P5
Then you heard wrong. WarpUp _disables_ the PowerUp kernel and makes
program developed for PowerUp impossible to run until you reboot next time.
>stops suporting the HAL....
Why would they since they created the hardware _and_ software from the
beginning.
Besides, a HAL is per definition something that is created by the hardware
and software developer, not by someone that has hacked the hardware.
>> > >Furthermore, phase 5 digital products will not tolerate that an
>> > >incompatible software takes over the control of the PowerUp hardware and
>> > >prevents the user from running software that has been developed for the
>> > >PowerUp boards by phase 5 digital products as well as many other
>> > >vendors.
>> > Now THIS is scary. I do not like that attitude.
>Yes this is scary. Will they tolerate Linux, NetBSD or other free
>software. Will they throw down anything that might be simmilar to
>their software? Do they want a monopoly?
They simply wants people to be able to use the board what they bought it
for. If someone wants to create a Linux or NetBSD for it, I doubt very much
that P5 would say anything about it since it does not create problems for
PowerUp developers/users.
When you on the other hand create software that makes compliant software
break, _that_ is scary.
>How do you know that they don't work together? Have you tested it?
>Since the H&P soft uses the HAL as well it should be no problem,
>shouldn't it?
No, it does not use the HAL.
>Can anyone test that? (Anyone not already in favour in one or the other)
Last time I tested WarpUp on my PowerUp development board it crashed right
away.
Randolph H. Murdock III
>Rudi Chiarito wrote:
>>
>> >Could it be that the demo showing the "advantages" has been provided by
>> >Haage&Partner?
>>
>> Of course, you can't expect a demo of WarpUp from anyone else but them ;)
>Yes, and the demo program running under the PowerUp System Software has
>also been provided by them. Oh, by the way, was it the "Set Pixel"-Demo
>which they have shown you? If so, this demo is as useful as testing a
>Porsche by driving backwards.
snippity snip snip
Much ado about the current WarpOS softs. I personally wish to put my two cents
in here. I do not plan on purchasing ANYTHING PPC until I see a native AOS for
it.
*** Team AMIGA ***
* <- Tribble o <- Jean Luc Tribble
Stefan Burstrom
On 01-Oct-97 11:14:04 Claude Mueller (cla...@hdclick.links.ch) wrote:
>Hallo w...@gf.phase5.de (Wolf Dietrich)
>Du hast am 30.09.1997 folgendes unter "Re: The Wolf is released! (was Re:
>WarpOS released !!!)" von Dir gegeben:
>> the advantages of their concept (and in my personal opinion, their pOS
>> is - even on 68k - still far from being a complete OS rather than a
>> replacement for some GUI and middleware parts of the Amiga OS)
>But let's say their pOS will be a complete OS someday. Why shouldn't it
>run on PowerUP ? That can only mean sellign more boards to you. You
>could have your software take care of the people that use AmigaOS,
>and proDad could write the drivers for their pOS for PowerUP.
>As for the Haage&Partner thing, I agree with you. But pOS is/will be
>a seperate OS, not to be compared with your approach.
Phase 5 never said that pOS would be banned from PowerUp. Phase 5 only said
that pOS wasn't their primary choice. If Phase5 thought that pOS shouldn't
run on PowerUp, why are ProDAD still PowerUp developers?
Stefan Burstrom
On 01-Oct-97 15:39:03 Steffen Haeuser (mag...@birdland.es.bawue.de) wrote:
> ˙° mo...@bulldozer.rydnet.lysator.liu.se schrieb : ˙´
>Hi!
mo>> No, no, no. Running WarpOS means that P5 software, or software created
mo>> with P5's _free_ devkit won't run anymore. Such a solution is a big step
mo>> backwards, back to the good old days when you had to reset the computer
mo>> between running programs.
>Only, if you use the WarpOS V8+. WarpOS V7 (a wrap-around on ppc.library)
>will always be available, and it runs parallel with the Phase 5 things.
The documentation for WarpOS sais that all functions in WarpOS V8+ will not
be available in the 'compatible' WarpOS V7. Too bad it isn't the user but
the software developer that chooses what technology the software is based
on.
Conclusion: There will always be programs that disables the real PowerUp
kernel just because some stupid developer though WarpOS was nifter at the
expens of incompatible software.
Stefan Burstrom
On 01-Oct-97 15:14:08 Steffen Haeuser (mag...@birdland.es.bawue.de) wrote:
>Well, now the problem. For example 68k task 1 is NOT independent of PPC task
>3. When they exchange data using the messaging system, a Cache Flush of the
>PPC is needed (else you will end in datalosses, i won't get into detail
>here... the docs of WarpOS and ppc.library explain this both). While this
>Cacheflush is done, all PPC Tasks come to a halt. Now use stated before your
>new messaging system would take "at worst" 5 microseconds for a message
>exchange. Let us calculate some numbers:
>the PPC has 32 KB Data Cache. 200000 times 5 microseconds would be a second.
>So we need to flush 32768*200000 Bytes per second. That would be:
>6.553.600.000 Bytes. This would mean, to make the time you announced you
>would need a memory interface that can do 6.5 GIGABYTE / second. I don't
>think PowerUP can do this. In fact i don't think ANY computer on this planet
>can do this.
Well, the lowend SGI O²'s with a X-bar bus does 1.2 GB/sec, so I don't
think it would be very hard to find a highend computer system that does
alot more;
>If there is a mistake in my calculations, please correct me. But i do not
>think so.
Yes, they are wery wrong and should really be redirected to dev/null.
Please get some facts before you post things like this.
>Now let us calculate the same with the statement of what Sam Jordan told
>about the WarpOS V8+ (he said on the PPC604e it would need 0.4-0.5 ms for a
>message).
>0.5 ms (worst case) would be 32768*2000 Bytes per second. This would mean we
>ended up with 65 MB/second, what sounds to me like a much more reasonable
>value.
The caches are not the big deal here.
A complete cacheflush on a copyback cache hardly flushes all cache lines.
Only the dirty lines are flushed. P5's API also features a 'no memory' API
where no memory areas are passed, only a set of LONG's. Then no caches
needs to be flushed.
The delay is caused by the time to switch from the current executing m68k
task to the m68k task that is the recipient of the PPC msg.
>>H&P have been promoting their Storm C with a concept that you simply
>>recompile the existing C code which has been written for 68k Amiga and
>>tell the compiler to generate PPC code. This may be very comfortable for
>>the developer, but is completely unefficient. Not only that in the worst
>Currently it is more efficient than ppc.library, prove your statements in own
> developpements before you state them !!! Also, it is, like outlined above,
>no streight recompile. Internally WarpUP works with tasks that talk to each
>other
So tell us then why StormC is faster. Do _you_ have the facts to back that
one up? I guess not!
>(so-called mirror tasks). But (as to what i outlined above) only one of them
>can be active at a time, the 68k one, or the PPC one.
>>case you have a random mix of 68k and PPC code, but also the two CPUs
><will not run in parallel as it is one continuous program flow which
>>jumps between PPC and 68k. If a linear code sequence is executed, the
>>68k and the PPC always work one after the other, in the worst case with
>>lots of switches and a useless overhead.
>Your statements about "parallel" still have to be proven. The calculations i
>did above seem to indicate that it is not possible.
If the program is rewritten to have 1 PPC task and 1 M68k task working at
the same time passing messages to eachother _without_ waiting for the
replies, you have true parallel execution. That is not the way StormC works
in the plain 'recompile and run' mode.
I can back this up with source codes and executables since I have been
developing for PowerUp for several weeks using GCC and P5's development
tools.
Jeroen T. Vermeulen
Summary of previous installment: As yet, very few Amiga software developers
have adapted their software to run on Phase 5's PowerPC hardware. To make it
easier for software developers to publish and maintain PowerPC software for the
Amiga, and to show what they think is possible with Phase 5's hardware, Haage &
Partners have released a free alternative OS for Phase 5's PowerUp boards.
Called WarpOS (or actually WarpUp if you include the API), this product lets a
developer simply recompile existing Amiga software to make some use of the PPC
processor's power. StormC has already been adapted to do this automatically.
Alternatively, software can also be ported by hand to make full use of the PPC
as is the case with Phase 5's solution.
The catch: You can't run software compiled for Phase 5's kernel on a board once
the WarpOS kernel is running. You need to reset your machine to switch between
the two kinds of software.
In response, Phase 5 throws a tantrum and threatens to modify their product to
prevent H&P's software from running. I don't speak for either company, but I'll
speak for myself just the same.
In article <342E63DF...@gf.phase5.de> Wolf Dietrich <w...@gf.phase5.de> writes:
>
> Furthermore, phase 5 digital products will not tolerate that an
> incompatible software takes over the control of the PowerUp hardware and
> prevents the user from running software that has been developed for the
> PowerUp boards by phase 5 digital products as well as many other
> vendors.
[ snip ]
> COMPLETE INCOMPATIBILITY MESS. In the interest of our customers, who buy
> our products with high expectations, and all our developers who expect a
> completely compatible product line, we will take care that the phase 5
> system software can not be disabled.
And where would Phase 5 have been today if Amiga Technologies had said the
same thing about CyberGraphX? Or Phase 5's 68060 library? Or Phase 5's FFS
patch? Right now I'm typing this on a Terminator-strength Amiga: Commodore
box outside, Phase 5 machinery inside. That would never have been possible
if these developments had been received by Commodore or AT like H&P's work
is now received by Phase 5.
Personally I have nothing against Phase 5, but this statement is clearly
inspired more by anger and frustration than by clear thinking. This is
Microsoft talk, it's written in Microsoft style (even referring to H&P as "a
german company" while the same goes for Phase 5 and most of their customers)
and with Microsoft arguments and using Microsoft words, and it helps nobody
but Microsoft.
For anybody who thinks I'm exaggerating, look up the discussion about the
proposed ISO standardization of Java. You'll see exactly the same
arguments. "Java is proprietary", although the specification is openly
available and anybody can implement it--just like H&P's extended hunk
format. "Java is optimized for Sun's own products" because it adheres to
accepted standards instead of being optimized for Intel--just like H&P's
solution supports the existing programming model and hunk format instead of
Phase 5's.
Personally I have trouble believing that this load of vitriol was written by
Wolf Dietrich who built up Phase 5 with his own vision, hard work, and
financial risk. I'd rather presume that it was submitted to him by one of
his own programmers who fears losing his position within the company as the
Tech Genius, the Amiga Guru with all development documentation inside his
head, if Phase 5 were to talk openly with people who did similar work better
in less time and give it away for free. We've seen this kind of internal
power play before, and in one case Phase 5 even had to publicly distance
themselves from claims made by some of their employees.
Something great might come out of a collaboration between Phase 5 and H&P on
this issue, especially for the Amiga as a whole, but not a brain monopoly
for Phase 5's programmer. OTOH perhaps Phase 5 could come up with arguments
that would convince H&P to conform to Phase 5's wishes, and they could talk
about that. But the single most stupid option would be to exclude any kind
of rational discussion right away by going hysterical.
> Haage&Partner has - from the beginning - rejected to use the ELF format
Is that why H&P's Website says WarpOS supports both ELF and the hunk format?
One can argue over these two companies' choices in this regard (Phase 5
thinks moving to Intel's standard is the obvious choice, H&P thinks
extending the Amiga standard is the obvious choice), but a fair statement
would at least have covered both sides of the argument. Why close people's
ears beforehand by coming up with unreasonable flames? That will ruin both
companies' reputations and split the Amiga community, but achieve nothing
else.
On the technical side, "supporting ELF" really means two different things:
Loading executables written in ELF format is a simple matter of patching or
rewriting LoadSeg(). Even if WarpOS doesn't support it, it should be an
easy thing to add. Hardly a point of discussion.
The other meaning, having your compiler and linker work with ELF object
files, is basically irrelevant to the way the system runs, but forcing all
compiler systems (and particularly their linkers) to support ELF would IMHO
be unreasonable. I see no reason why an OS can't support both, and
different development environments can choose between the two alternative
formats. Both formats start with magic numbers, and both have their hunks
loaded and relocations performed in the same kinds of memory.
> Haage&Partner, instead, wants to establish their own "extended
> Hunk-format", which may is derived from the Amiga Hunk format, but still
> a new and proprietary solution.
Is that why it is openly documented? Is "new" the same as "proprietary"?
In what way is the EHF more proprietary than, say, Phase 5's picture
datatype which may not even be used with CyberGraphX clones such as
Picasso96? What is the problem with a "new" extension to existing formats
as compared to forcing developers to design completely new software?
Sure, there may be a real argument under this stab. But this way of
expressing it is not designed to convince anybody--it is designed to make
them stab back at Phase 5's own sore spots. The goal: Have everybody
fighting instead of working, and ultimately have us all wondering where yet
another year went that could have brought useful PPC Amigas to our desktops.
> It is a concept that forces developers
> to use the only solution supporting this format - the Storm C compiler
> by Haage&Partner. Haage&Partner declares it's product being open and
> Amiga-compliant - in fact, it is only open to the Haage&Partner standard
> and compliant to their own software solutions.
That's a strange argument: Source-level compatibility with existing
software, which WarpUp promises, is hard to argue with. What makes software
that must be written from scratch to Phase 5's own programming rules more
Amiga-compliant than an existing program that is just recompiled from source?
Besides, it also isn't true. I just looked at the WarpUp documentation and
it's extensive and open. It's some of the clearest, most concise
programming documentation I've ever seen. Nothing like what you get for
some of Phase 5's software: "This product does roughly the same as Enforcer
[no version specified] plus the following options, and I'm too lazy to write
documentation so you figure it out", "This may work on A4000: Try it, it
will work or not". Of course I can't compare the WarpOS docs to Phase 5's
PowerUp documentation because as a registered copper-level Phase 5
developer, I can't even *get* any PowerUp development documentation!
Yeah, sure, I'll get it when I buy the board. I can't just download to see
if it's worthwhile trying to prepare my software to run on Phase 5's PowerUp
system, or how to do so, or whether I can try to port it at all. Buy the
board first, then find out what you got. How's that for open and
Amiga-compliant? And when I buy the board, apparently I can't get any
updates from Phase 5's support site, whether under "technical support" or
under "developer support", because the files just aren't there.
If the public information about WarpUp is anywhere near correct, it has the
potential of getting thousands of applications running at PPC speed within
a few months, followed by full-power native applications at a slower pace.
We don't see that happening with Phase 5's system, so they should at least
be interested in talking about WarpUp. Anybody with a StormC compiler could
download sources from Aminet and recompile them for PowerPC; other compilers
could also be adapted, and apparently with greater ease than with Phase 5's
system because there's no need to switch to ELF.
Phase 5's solution would force you to rewrite the software *and* adapt the
compiler. That means higher costs, higher risks, and more effort for the
developer, resulting in less software. Thus WarpUp adds a significant
incentive for people to buy Phase 5's PowerUp cards, and we're being told
it's a Bad Thing. Somebody at Phase 5, it seems, is trying to "snatch
defeat from the jaws of victory".
> At this point, it is important to mention that the software concepts of
> phase 5 digital products have all the time been focused on developments
> that open doors and smooth upgrade paths to future and completely
> revised OS versions, which incorporate functionalities such as memory
> protection and multi-processing (which is not supported by the
> Haage&Partner concept, BTW).
Wolf, I think you've been misinformed. The story you told about WarpUp's
multiprocessing limitations and memory model was a misrepresentation of what
is in the WarpOS manual.
The process you describe where software must be rewritten to the PPC's
rules, and restructured so that a PPC task delegates work that must be done
by the "master" processor back to the 68k, is necessarily also there in
WarpOS. But WarpUp can simplify this process by automating it and inserting
the necessary cross-processor stubs for library calls. That means you can
get a quick, easy, relatively inefficient, but still clean port of an
existing source.
The resulting program "flips back and forth" between 68k and PPC subthreads
as needed. That doesn't mean that you lose the option of writing an
efficient new dual-processor version by hand, or of letting the PPC part and
the 68k part of the program execute simultaneously. Look up the
documentation for the RunPPC() and Run68K() context switch calls--see that
flag labeled PPF_ASYNC? The truth is that WarpOS looks a lot like having a
second exec running on the PowerPC, except that it doesn't let you Disable()
or Forbid() and properly protects shared data structures with semaphores.
> This has e.g. also been proposed to Amiga
> International/Gateway2000 in May this year, together with our offer for
> extensive technological cooperation. To reach such a goal, we feel it is
> mandatory to set today guidelines for programming that make current
> Amiga/PowerUp developments more independent from the current OS revision
> by clear structured and object-oriented programming, instead of using
> the StormC concept of simply recompiling older sources (this issue will
> be adressed in detail later).
Let's look at this argument: H&P should adhere to Phase 5's programming
rules, and forget the existing ones that define the Amiga OS, just because
"it is mandatory to set rules"? Why Phase 5's rules? Because they have
"been proposed" to Amiga International? All that that statement means is
that AI is aware of Phase 5's ideas, and has NOT endorsed them as official
guidelines. We might as well say that StormC has the first rights because it
existed before PowerUp!
What this appears to mean is that Phase 5, or at least somebody working
there, is trying to force Amiga programmers to move to a Phase 5-controlled
programming environment (presumably the A\Box OS), by forcing us to adhere
to its rules right now. The PowerPC is an excellent opportunity for that:
Couple the 680x0->PPC transition to the AmigaOS->Phase5 transition so
developers have no choice but to stick with Phase 5's system, so then users
have no choice. And when AInt/AInc create an official PowerAmiga they are
faced with a fait accompli: They will have to stick to Phase 5's solution
because all existing PPC software is already using it. And since Phase 5's
solution is planned to be compatible with the A\Box, they would have all the
power from then on and the Amiga would have to follow them instead of the
other way around.
All this, of course, at the cost of having almost no software available
right now that makes use of the PowerUp board. And once the software is
there, this strategy ensures that "compatible" Amiga software can use no
significant features that the A\Box doesn't have. Presto: Total industry
takeover. Personally I wouldn't mind seeing Phase 5 steering the Amiga
platform towards the future, but whoever is responsible for this fanatical
reaction against WarpUp will have to go first.
We've seen the same kind of thing when some of the people involved here
violently attacked AT's New Style Devices standard: They said that their
"competing" trackdisk64 standard was the official one (the two could
actually live together fine but the TD64 authors used it as an argument
against NSD anyway, then denied that the possibility of giving TD64 its own
official place within NSD was ever discussed, and even cancelled somebody
else's Usenet post that proved them wrong) because several third-party
developers were involved, whereas they stubbornly called Amiga Technologies'
NSD standard a "draft". But when I looked at two different public-release
versions of the TD64 spec, they were both actually labeled "preliminary"
themselves.
But here's what I see as the possible outcome of frank, constructive
discussion over these two PowerUp systems:
Phase 5's statement says that their solution is much more future-proof and
OS-independent, enforces more structured and object-oriented programming
styles, and is prepared for more advanced features w.r.t. multiprocessing
and memory protection. Then DECOUPLE the 680x0->PPC migration from the
AmigaOS->Phase5 migration by working together to implement the Phase 5 API
atop/besides the WarpUp API (the latter is actually very neatly decoupled
from the WarpOS operating kernel to ensure hardware independence), or at
least a system derived from WarpUp with adaptations to enable the two to
live together well.
Then *recommend* that all new software be written to Phase 5's allegedly
stricter, cleaner rules and API for future compatibility. Those new rules
can then be considered by AInt/AInc and hopefully endorsed as official
guidelines. It can be done now, while there is essentially no existing
software to prevent such adaptations.
I think the past has shown that the Amiga programmer community, especially
since the release of KickStart V36, is willing to go the extra mile to make
their work live up to higher internal quality standards. If Phase 5 can
convince them that their way is better for the future, they will find a
ready ear. But there is also no doubt that we NEED SOURCE-LEVEL PPC
COMPATIBILITY NOW. From what I've seen of WarpUp, I think it can achieve
this level of backwards-compatibility with a minimum of compromise.
OTOH, just adding stricter programming rules for "future compatibility" can
also be risky. Not only because they may not match exactly the Amiga's
ultimate development direction, but also because some of those "sloppy"
programming practices and software constructs that make memory protection so
difficult on this system--our shared libraries, OS patching, messaging,
public memory--may eventually turn out to be the real strengths of the
Amiga. We can't afford to squander those elusive but vital qualities in
return for a memory-protected future that we may never quite reach anyway
because it would break all existing software!
Other systems are based on solid, straightforward ideas of reliability and
protection. That's good, but the Amiga has something different: Removal of
restrictions rather than addition of isolated features. Orthogonality of
components. Flexibility. Different tradeoffs. Its own view of basically
everything. As a result, the system missed out on some clearly defined
features but I like to think it's also gained something. It may be hard to
put one's finger on it because it isn't on anybody's feature list, but it
could also turn out to be very easy to destroy by turning the platform into
yet another "off-the-shelf" platform that tries to be everything to
everybody who already uses Windows or UNIX.
So Phase 5, stop shouting and start talking before we go into another
year-long flamewar that ends in a stalemate! If you decide to slug this one
out, the time spent fighting and the opportunity to advance the platform
will be wasted and your good reputation will disappear no matter what the
outcome.
> As the solution from Haage&Partner is not compatible and can - because
> of it's proprietary software concept - not fullfill demands of
> professional developers, it is very unlikely that developers will choose
> it.
Haage & Partners have made it very clear that WarpUp will work with other
compilers as well as StormC, but that StormC has been extended with the
ability to automatically "port" software to WarpUp. From what I've seen,
other compilers could also be adapted to make use of this capability,
probably by writing new pragma/inline headers for the PPC code. PPC
software written for WarpOS must comply to stricter rules than have so far
been usual for the AmigaOS, as is the case with Phase 5's system, but it's
done in such a way that existing Amiga sources can be recompiled to
automatically adhere to these rules and still run by delegating the "nasty"
work to the 680x0 side.
Risk-free easy porting--does that not fulfill the demands of developers?
The only apparent reason for making this prediction is to scare people away
from WarpOS by making them believe that nobody else will use it. If Phase 5
has a good technical argument, why resort to intimidation?
> phase 5 digital products has supported Haage&partner with free developer
> systems, as well as all necessary information to develop compatible
> software and tools, thus enabling Haage&Partner to become a leading
> software vendor.
This is nonsense. Either Phase 5's products are compatible to the Amiga
system so their developer support was irrelevant for H&P's developing
"compatible" software and "becoming a leading software vendor", or Phase 5's
products aren't compatible and H&P needed support to make things run on
them, but then everything that is said here about compatibility is untrue.
Or should that be, a leading *PowerUp* software vendor? Apparently not,
because according to this same statement...
> At the same time, they have yet not been able to present a stable
> running PowerPC application which the Amiga/PowerUp user would have an
> advantage from.
In that case, who has? I know of two 2D graphics manipulation packages and
a renderer, as well as a compiler. Two of those are by H&P, so I guess they
aren't "stable running PowerPC applications". But at least H&P have worked
to create an ability to recompile Amiga software for the PPC processor in
the PowerUp boards. That's exactly what we've all been waiting for, yet
Phase 5 now says it is against our interests. At the very least they should
be reconsidering their own work, and asking themselves if WarpOS hasn't
shown them that they can do better. I know their work is usually of the
highest quality, but like all of us they should face the possibility that
you can't out-think everybody else all the time.
> Haage&Partner, and they even attempt to trash the complete and
> comprehensive software development in which we have spend many efforts.
Personally, I'd like to know more about this extensive software development.
Where is it? Supposedly this is not just the PowerUp operating software,
because three H&P programmers did more or less the same thing in their spare
time without documentation. It also can't be those promised PPC-native
CyberGraphX extensions, because the standards set in this statement make
them "closed, proprietary software concepts".
But that's not really relevant. Here's what is: If all those other PowerUp
applications exist but simply haven't been released yet, it remains to be
seen whether the effort would really go to waste if Phase 5 and H&P were to
work together on a unified PowerPC solution--especially if it can support
both APIs. Maybe some porting work will have become unnecessary when
source-level Amiga compatibility is attained, but what that really means is
that developers have been jumping through unnecessary hoops. The only
actual risk is increased competition for ported applications and for the
A\Box as WarpOS makes more Amiga software available for those PPC boards.
> We highly disregard this contraproductive and
> confrontational behaviour, and will completely drop any support for
> Haage&Partner as a PowerUp developer as well as for any of their
> products.
And there we are: Yet another crack in the Amiga platform, with Phase 5
criticizing the same thing that they themselves have done in the past.
Personally I consider Phase 5 a top-notch technology company and a major
driving force behind the continued development of the Amiga, but they must
rid themselves of the destructive influences that have caused so much untold
damage. My hardware supplier had to give up dealing with Phase 5 directly
because of the treatment he got from them. So far I have found no PowerUp
developer support and documentation for the shareware/freeware programmers
that are absolutely essential for the survival of our platform. Other
developers have seen their efforts ridiculed and discouraged by certain
Phase 5 people, and now we see developer support used as a weapon to wage
war on competitors. That is unacceptable if either Phase 5 or the Amiga is
to do well in the future.
--
,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,
; Jeroen T. Vermeulen \\"How are we doing?"// Yes, we use Amigas ;
;--- j...@xs4all.nl ---\\"Same as always."//-- ... --;
;jver...@wi.leidenuniv.nl \\"That bad huh?"// Got a problem with that? ;
''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
If Bill Gates wants you dead you gotta be doing something right
Justin Smith
Stefan Burstrom <f94...@efd.lth.se> wrote in article
<1276.7214...@efd.lth.se>...
> The documentation for WarpOS sais that all functions in WarpOS V8+ will
not
> be available in the 'compatible' WarpOS V7. Too bad it isn't the user but
> the software developer that chooses what technology the software is based
> on.
> Conclusion: There will always be programs that disables the real PowerUp
> kernel just because some stupid developer though WarpOS was nifter at the
> expens of incompatible software.
I can understand why phase5 is pissed. They are delivering the PowerUp
boards with the promise that software developed for them (using the PowerUp
software) will run. Using WarpUp, H&P break this promise between phase5
and purchasers of PowerUp boards and software.
Justin Smith
University of New Mexico
jths...@unm.edu
Saving for CyberstormPPC
Dreaming the A\Box dream
James McArthur
Steven Matty (Big.Wil...@ldngedge.demon.co.uk) wrote:
[ .. ]
: The way I see it - it's a thinly disguised attempt to force
: Amiga developers to use the StormC compiler as "The Standard"
: which IMHO is a VERY BAD THING INDEED(tm)
: I hope the majority of developers see sense and refuse to
: fall for it.
And at almost $700 AUS for the Storm C compiler, I personally wont be using
it.. esp when the GNU utilities are free and generally of better quality IMO.
--
Amiga 4000/060/PPC150 CV64/3D 14Mb James McArthur
Currently reposting ALL spam to:
*@iemmc.org *@answerme.com *@global-impact.com *@cafes.net
*@mail.enterprise.net
James McArthur
Wolf Dietrich (w...@gf.phase5.de) wrote:
[ .. ]
: Once again, we ONLY want to make sure that customers of our products can
: use software based on the PowerUp System Software and the PPC library at
: any time. If this means that WarpOS V8+ may not run, it is a consequence
: of the design that H&P have been choosen.
So a WarpOS that maintained compatibility with the ppc.library, and added
extra functionality would be acceptable to P5?
kno...@lsil.com
In article <60qjn8$t...@arcadia.informatik.uni-muenchen.de>,
ort...@informatik.tu-muenchen.de (Mathias Ortmann) wrote:
>
> mo...@bulldozer.rydnet.lysator.liu.se (Mans Engman) wrote:
>
> >Yes, but on the other hand I don't want to see a fragmentation right
> >now. Run P5 software -> reboot -> run warpos software -> reboot...
>
> Why is frequent rebooting only a problem when it happens on the powerUP?
>
> When the PIOS ONE was still an issue, nobody complained about having to
> reboot to switch between applications available for MacOS only, Linux-
> specific programs and BeOS-specific software (is there any?).
You seem to be confusing two totally seperate issues here Mathias.
On the one hand, the PIOS would appear to require rebooting in order to
switch between different OSes.
On the other hand, thanks to Haag&p's attempts to secure a larger market
for StormC via sneaky backdoor tricks, they will require us to reboot to
run different software UNDER THE SAME OS (AmigaOS).
Now, I don't know about you, but I personally find it a huge leap
backwards if I will be required to reboot my Amiga to run different Amiga
apps.
H&P have just taken the first step towards fragmenting an already small
market, purely for their own greed - a VERY bad move indeed.
Just hmo, of course.
TTFN,
Keith
-------------------==== Posted via Deja News ====-----------------------
http://www.dejanews.com/ Search, Read, Post to Usenet
Wolf Dietrich
James McArthur wrote:
>
> : Once again, we ONLY want to make sure that customers of our products can
> : use software based on the PowerUp System Software and the PPC library at
> : any time. If this means that WarpOS V8+ may not run, it is a consequence
> : of the design that H&P have been choosen.
>
> So a WarpOS that maintained compatibility with the ppc.library, and added
> extra functionality would be acceptable to P5?
Exactly one week ago I had a two hour meeting with Mr. Haage from
Haage&Partner, where he picked up some more Cyberstorm PPC boards. He
informed us about there current status of development of Storm C abd
their compatible powerpc.library, and - although we follow different
approaches - we were in perfect agreement that a compatible alternative
which is completely transparent for the users is no problem at all, and
that developers may decide for themselves which development approach is
the right one for them. Later then, while showing us some of their
demos, Mr. Haage surprisingly admitted that they now will push another
solution which is NOT compatible and disables all the system software we
provide and everyhting that runs on top of it, which they apparently
have been working on for quite a long time.
So, to answer your question: Yes. Everything that supports PowerUp is
appreciated from our side, as long as it does not disable functionality
that we provide and that our customers and developers want to use. We do
also support developments such as Linux for the PowerUp, which is a
completely different solution.
Alain Penders vh3
Jeroen T. Vermeulen (j...@xs4all.nl) wrote:
: The catch: You can't run software compiled for Phase 5's kernel on a board once
: the WarpOS kernel is running. You need to reset your machine to switch between
: the two kinds of software.
:
: In response, Phase 5 throws a tantrum and threatens to modify their product to
: prevent H&P's software from running. I don't speak for either company, but I'll
: speak for myself just the same.
Ofcourse, we all know that you're not the project manager for Merapi, that
you don't work for them and that you couldn't possibly be biased by that.
: And where would Phase 5 have been today if Amiga Technologies had said the
: same thing about CyberGraphX? Or Phase 5's 68060 library? Or Phase 5's FFS
: patch? Right now I'm typing this on a Terminator-strength Amiga: Commodore
: box outside, Phase 5 machinery inside. That would never have been possible
: if these developments had been received by Commodore or AT like H&P's work
: is now received by Phase 5.
Those things were developed at times were C=/AT/AI wasn't working on the
Amiga and could not possibly have developed them. Phase5 is actively developing
on the PowerUP boards and A\Box. So no, this doesn't compare.
: format. "Java is optimized for Sun's own products" because it adheres to
: accepted standards instead of being optimized for Intel--just like H&P's
: solution supports the existing programming model and hunk format instead of
: Phase 5's.
Doesn't it ? SAS/C does ELF... GCC does ELF... none of the other compilers
do either ELF or H&P's format. As far as I can see, the H&P format is
*solely* used by Storm C++.
: about that. But the single most stupid option would be to exclude any kind
: of rational discussion right away by going hysterical.
They developed a kernel that they're going to base their OS on. A 3 year
old child can understand that an external company trying to kick that kernel
is asking to be kicked back.
: > Haage&Partner has - from the beginning - rejected to use the ELF format
:
: Is that why H&P's Website says WarpOS supports both ELF and the hunk format?
Is that why Jochen Becker in his recent reply to Phase5 stated that WarpUP
does NOT support ELF ?
: One can argue over these two companies' choices in this regard (Phase 5
: thinks moving to Intel's standard is the obvious choice, H&P thinks
ELF isn't an Intel standard, and you know it.
: On the technical side, "supporting ELF" really means two different things:
: Loading executables written in ELF format is a simple matter of patching or
: rewriting LoadSeg(). Even if WarpOS doesn't support it, it should be an
: easy thing to add. Hardly a point of discussion.
Then why does Jochen Becker state in his reply that it is ? I guess his
statement went up in the air after it became public that Steve Krueger
needed *3* days to add it to SAS/C... on his own... in his spare time.
: > It is a concept that forces developers
: > to use the only solution supporting this format - the Storm C compiler
: > by Haage&Partner. Haage&Partner declares it's product being open and
: > Amiga-compliant - in fact, it is only open to the Haage&Partner standard
: > and compliant to their own software solutions.
:
: That's a strange argument: Source-level compatibility with existing
: software, which WarpUp promises, is hard to argue with. What makes software
: that must be written from scratch to Phase 5's own programming rules more
: Amiga-compliant than an existing program that is just recompiled from source?
The Storm solution is a mega cludge. Amiga is in need of new things... a more
modern OS. The most logical place to break backward compatibility and fix
things is when changing the CPU. (Since programs have to be recompiled at
that time, it's only logical that their programmers are still active and can
actually change the code.) That's what Phase5 is trying to do. H&P's
solution contributes to short term profit... allowing ppl to instantly
port their SW to PPC, without any effort at all. The resulting system is a
cludge, and it'll be alot harder at that point to convince people to port
their programs to a decent non-compatible architecture.
: If the public information about WarpUp is anywhere near correct, it has the
: potential of getting thousands of applications running at PPC speed within
: a few months, followed by full-power native applications at a slower pace.
: We don't see that happening with Phase 5's system, so they should at least
: be interested in talking about WarpUp. Anybody with a StormC compiler could
: download sources from Aminet and recompile them for PowerPC; other compilers
: could also be adapted, and apparently with greater ease than with Phase 5's
: system because there's no need to switch to ELF.
Does the documentation give information on what code a compiler has to
produce to be able to do H&P's "fast" context switch ? How to do the
part 68k / part PPC code in one executable ? If not, then there is no
technological advantage to be gained from WarpUp by any other compiler
manufacturer.
: Phase 5's solution would force you to rewrite the software *and* adapt the
: compiler. That means higher costs, higher risks, and more effort for the
: developer, resulting in less software. Thus WarpUp adds a significant
: incentive for people to buy Phase 5's PowerUp cards, and we're being told
: it's a Bad Thing. Somebody at Phase 5, it seems, is trying to "snatch
: defeat from the jaws of victory".
Ofcourse, asking Jim Drew to make an 68k Amiga emulator on top of PowerUP
would be an even better solution than WarpUP.
: Let's look at this argument: H&P should adhere to Phase 5's programming
: rules, and forget the existing ones that define the Amiga OS, just because
: "it is mandatory to set rules"? Why Phase 5's rules? Because they have
: "been proposed" to Amiga International? All that that statement means is
: that AI is aware of Phase 5's ideas, and has NOT endorsed them as official
: guidelines. We might as well say that StormC has the first rights because it
: existed before PowerUp!
Seems like that is how H&P thinks. AI hasn't announced any concrete plans
for a new Amiga... yet Jochen Becker acted as if an upcoming PPC Amiga is
a sure thing, and that you *have* to use WarpUP to be compatible with it.
: What this appears to mean is that Phase 5, or at least somebody working
: there, is trying to force Amiga programmers to move to a Phase 5-controlled
: programming environment (presumably the A\Box OS), by forcing us to adhere
: to its rules right now. The PowerPC is an excellent opportunity for that:
: Couple the 680x0->PPC transition to the AmigaOS->Phase5 transition so
: developers have no choice but to stick with Phase 5's system, so then users
: have no choice. And when AInt/AInc create an official PowerAmiga they are
: faced with a fait accompli: They will have to stick to Phase 5's solution
: because all existing PPC software is already using it. And since Phase 5's
: solution is planned to be compatible with the A\Box, they would have all the
: power from then on and the Amiga would have to follow them instead of the
: other way around.
Maybe that's true. Would be smart marketing. It's their hardware... so who
is H&P to think they can play god and stop P5 ? I guess they already found
out how well that works.
: All this, of course, at the cost of having almost no software available
: right now that makes use of the PowerUp board. And once the software is
: there, this strategy ensures that "compatible" Amiga software can use no
: significant features that the A\Box doesn't have. Presto: Total industry
: takeover. Personally I wouldn't mind seeing Phase 5 steering the Amiga
: platform towards the future, but whoever is responsible for this fanatical
: reaction against WarpUp will have to go first.
No software means no/less sales. Why do you or H&P care if a move P5 makes
ruins their sales ? And how can such a move *contribute* to industry
takeover?
: We've seen the same kind of thing when some of the people involved here
: violently attacked AT's New Style Devices standard: They said that their
: "competing" trackdisk64 standard was the official one (the two could
: actually live together fine but the TD64 authors used it as an argument
: against NSD anyway, then denied that the possibility of giving TD64 its own
: official place within NSD was ever discussed, and even cancelled somebody
: else's Usenet post that proved them wrong) because several third-party
: developers were involved, whereas they stubbornly called Amiga Technologies'
: NSD standard a "draft". But when I looked at two different public-release
: versions of the TD64 spec, they were both actually labeled "preliminary"
: themselves.
Hardly P5 related. P5 was one of many 3rd party developers who supported
TD64 and had a problem with NSD. Most of them still have a problem with
it. I think there are more TD64 compatible scsi controllers out there
then there are NSD compatible ones.
Alain
--
Alain Penders | Visserijstraat 114 | http://www.Finale-Dev.com/
Finale Development, Inc. | B-3590 Diepenbeek | Finale Web Cruiser
Al...@Finale-Dev.com | Belgium | MOca Java Virtual Machine
DISCLAIMER: The contents of this mail reflect my own opinions, and
do NOT in any way represent Finale Development, the ICOA or the TSC.
Espen Berntsen
On 01 Oct 1997 14:41:09, mag...@birdland.es.bawue.de (Steffen
Ah, so they won't be making some stupid hardware key to prevent
software from running on the CPU board (talk about smilminded,
spitefull ..guys.. if they had done this).
--
Nameless <> Name...@mo.himolde.no
Let's face it, if there is one thing that's addictive: it's politics. And
politics, unlike pot, most definitely _does_ cause brain damage.
Michael Rock
Stefan Burstrom wrote:
>
> On 30-Sep-97 21:21:21 Goswin Brederlow (goswin.b...@student.uni-tuebingen.de) wrote:
> >Wolf Dietrich <w...@gf.phase5.de> writes:
>
> >> George Noel wrote:
>
> >It doesn't use the HAL? I heard something very different. And if P5
> Then you heard wrong. WarpUp _disables_ the PowerUp kernel and makes
> program developed for PowerUp impossible to run until you reboot next time.
Ok, for another time :
WarpOS itself sits _under_ the powerpc.library.
There are two versions of powerpc.library, one that uses WarpOS as a HAL
layer,
and another one called V7 that uses ppc.library as a HAL layer.
From the top view, both look the same, only the V7 is lacking some
features
that WarpOS provides. The bebefits from this is that a programm written
for powerpc.library can be used on both systems. You as a user may
decide to
install WarpOS on your system if you think you can go without any ELF
codes,
but have some applications that do better on WarpOS.
Or you can install V7 which will go well on ppc.library by P5, as long
as they
do not stop it from doing so.
> >stops suporting the HAL....
>
> Why would they since they created the hardware _and_ software from the
> beginning.
>
> Besides, a HAL is per definition something that is created by the hardware
> and software developer, not by someone that has hacked the hardware.
>
>> > >prevents the user from running software that has been developed
for the
> >> > >PowerUp boards by phase 5 digital products as well as many other
> >> > >vendors.
>
> >> > Now THIS is scary. I do not like that attitude.
>
> >Yes this is scary. Will they tolerate Linux, NetBSD or other free
> >software. Will they throw down anything that might be simmilar to
> >their software? Do they want a monopoly?
>
> They simply wants people to be able to use the board what they bought it
> for. If someone wants to create a Linux or NetBSD for it, I doubt very much
> that P5 would say anything about it since it does not create problems for
> PowerUp developers/users.
> When you on the other hand create software that makes compliant software
> break, _that_ is scary.
It only breakes IF YOU as the user decide to use it and tell it to do
so.
> >How do you know that they don't work together? Have you tested it?
> >Since the H&P soft uses the HAL as well it should be no problem,
> >shouldn't it?
>
> No, it does not use the HAL.
It can, if you want it to. But in that case you will have to live with
the
ppc.library under it and need to forget about things like
a fast memory-map setup for your games, because ppc.library does not
allow
for this. (up to now, but who knows what will happen to ppc.library)
> >Can anyone test that? (Anyone not already in favour in one or the other)
>
> Last time I tested WarpUp on my PowerUp development board it crashed right
> away.
Ok, you can always call the H&P hotline about that.
Maybe you had a ppc.library already running ?
You may rename that library to make sure it does not pop up in the wrong
moment.
> regards, Stefan Burstroem
>
> -----------------------------------------------------------------------
> >> Irl: Stefan Burstroem << >> Omnipresence Intl. << >> Irc: Yabba <<
> >> Phone: +46 (0)46-211 40 84 << >> EMail: ste...@omnipresence.com <<
> -----------------------------------------------------------------------
--
Michael Rock, student of computer science, TU Braunschweig.
Also at
Haage & Partner PPC-RISC Compiler department.
contact me at
m.r...@haage-partner.com
A3oooT 060@50 + 603e@150
Alan L.M. Buxey
On 30 Sep 1997 04:57:47 GMT ,George Noel posted the following:
: >Haage&Partner, instead, wants to establish their own "extended
: >Hunk-format", which may is derived from the Amiga Hunk format, but still
: >to use the only solution supporting this format - the Storm C compiler
: >by Haage&Partner.
: But instead, you also want developers to ONLY use your solution so what
: makes you any better in that regard?
dont be silly - this isnt the case at all..
i should be able to compile stuff for PowerPC no matter WHAT compiler
i'm using - like Phase5 have managed to achieve - eg using gcc (and in
future SAS/C)
using some hack to access the PowerPC chip is not the way to go - esp.
if you have to reset to be able to use any other PowerPc program using
ppc.library - eg all of Phase5's stuff
alan
kno...@lsil.com
In article <1703.7213...@ldngedge.demon.co.uk>,
Steven Matty <Big.Wil...@ldngedge.demon.co.uk> wrote:
>
> I've been reading this thread with interest, and thought I'd add
> my bit...
>
> Good Things(tm) about WarpOS
> ----------------------------
>
> 1) It is (allegedly) faster
>
> Bad Things(tm) about WarpOS
> ---------------------------
>
> 1) Programmers wishing to write for it *MUST* use H&P's commercial
> C compiler
>
> 2) It is completely incompatible with the DE-FACTO PowerUP kernel
>
> 3) It won't be possible to write code with GCC or SAS/C and run
> these programs nuder WarpOS (afaik, WarpOS will not run programs
> compiled to use Phase5's library in ELF format)
>
> 4) Having to reboot to run a PowerUP program or a WarpUP program
> is just plain awful
>
> 5) It's a kludge - no matter how you present it, it is based on
> reverse engineering of the PowerUP hardware - and is more
> likely to break on future (or current?) cards.
>
> The way I see it - it's a thinly disguised attempt to force
> Amiga developers to use the StormC compiler as "The Standard"
> which IMHO is a VERY BAD THING INDEED(tm)
Exactly. Very well said.
I for one will be exceedingly pissed off with H&P if, as is likely, it
turns out that if I want to run some programs written using the STANDARD
ppc.library, then something which someone wrote using ht eH&P PROPRIETARY
system, I have to reboot, or cannot run them both in parallel.
As you rightly point out, this is purely a cynical ploy by H&P to try to
force everyone to use their compiler, thereby eliminating the competition
(the free gcc and, very soon, the free SAS/C PPC upgrade).
I certianly hope that develoers and users alike see H&Ps attempts to take
over teh market for what they are, act act accordingly.
(Of course, it is interesting to remember that code can be written for
the P5 ppc.library system under 2, soon 3, compilers, whereas teh H&P
system FORCES you to use, suprise surprise, teh H&P commercial compiler
only.)
TTFN,
Keith
--
Michael Rock
Hello #?.
Per Jonsson wrote:
> But teher is no response or reaction from H&P, if I got a better
> product and got a lot of "shit" in a newsgroup I would scream.
> I want to see some response from some at the H&P base.
Statements are currently being written and translated.
Nobody here expected this kind of over-reaction from P5.
> > What a waste of time, energy and probably good programmers, they should
> > have done something positive for the Amiga.
To speak for me, there was no waste of energy and time.
And i have the strong feeling we did someting positive for the
amiga, but the whole story is a bit to complex to be written down here.
> It's a pity. Instead of making WarpOS they could have joined P5 and
> maybe optimized the CyberGL just a little bit more.
What do you mean by "join" ?
Ask for employment ?
> I still hope this just is a big missunderstanding. I think it would be
> the best thing.
It looks like a combinition of misunderstanding combined with
over-reaction to me. But of course, i can only speak for myself.
> /Per Jonsson
M_R
Alan L.M. Buxey
On Wed, 01 Oct 1997 10:10:45 GMT ,Mathias Ortmann posted the following:
: Stop making yourself look like a fool. You know _exactly_ what the primary
: justification for the existence of P96 is: It has a _documented_ driver API,
: allowing independent third parties to develop P96 drivers for their gfx
stop talking drivel. if a 3rd party wants a CyberGfx driver for their
board, they can send a board plus some fees to the CGfx people to
develop one.
this actually works out a lot cheaper and faster than them trying to do
it themselves, gets a full, proper driver AND means that people may now
buy their board since it has CGfx support - well worth doing
: into just another building block for the micromonopoly you are trying to
: establish in the Amiga market.
shut up and go back to UAE
alan
Michael Rock
Justin Smith wrote:
>
> Stefan Burstrom <f94...@efd.lth.se> wrote in article
> <1276.7214...@efd.lth.se>...
> > kernel just because some stupid developer though WarpOS was nifter at the
> > expens of incompatible software.
... or that the deveoloper needs some low-level stuff that the powerup
does not
provide for, upto now.
> I can understand why phase5 is pissed. They are delivering the PowerUp
> boards with the promise that software developed for them (using the PowerUp
> software) will run. Using WarpUp, H&P break this promise between phase5
> and purchasers of PowerUp boards and software.
The V7 uses their ppc.library, so it will not break any promise.
But if you install the V12, that is your decision.
So how can you say that the ones who did the software you installed is
the
one to break any promise ?
> Justin Smith
> University of New Mexico
> jths...@unm.edu
> Saving for CyberstormPPC
> Dreaming the A\Box dream
--
Ralph Schmidt
written by
>Wolf Dietrich who built up Phase 5 with his own vision, hard work,
and
>financial risk. I'd rather presume that it was submitted to him by
one of
>his own programmers who fears losing his position within the company
as the
>Tech Genius, the Amiga Guru with all development documentation inside
his
>head, if Phase 5 were to talk openly with people who did similar work
better
>in less time and give it away for free. We've seen this kind of
internal
>power play before, and in one case Phase 5 even had to publicly
distance
>themselves from claims made by some of their employees.
>
Ok..Mr Vermeulen. You obviously badmouth me here and i can愒 let
this stand the way it is.
1) Wolf Dietrich is the marketing/business director of Phase5
which he owns by 50%
2) Gerald Carda is the technical(HW and SW) director of Phase5
which he owns by 50%
3) I was informed about the incident Friday evening after Gerald
and Wolf already decided how their reaction to this war declaration
by H&P will be.
Wolf and Gerald aren愒 the people which I殼 really able to
influence the way you want the people believe here.
Everybody who knows both knows that i say the truth here.
H&P told Phase5 on a *thursday* meeting after some smalltalk..
Oh..by the way we惻l release our own Kernel now which will be
incompatible with yours. Wolf and Gerald told them that this won愒
be accepted but Haage didn愒 seem to be interested into what they
told him..well..the whole history of events between Phase5 and H&P
showed this.
4) The Kernel *design* is a combined effort by Gerald Karda,
Frank Mariak, Andre Osterhues, me and the feedback of
developers which actually *cared* for it.
H&P didn愒 care for it...I got *1* email by this Mr. Jordan
which is Haage愀 PPC programmer over the last 10 months
about the PowerUP board. In this email(Sorry that i quote
from an email which is normally not my style) he demanded
a Forbid() in the PPC Kernel. I told him that i won愒
implement a Forbid() on an user level but will provide
semaphores and which is what i exactly did.
The EMail was from the 22th April 1997.
From Haage itself I got 1 bugreport which i fixed after a
1-2 weeks delay(they phoned me about this in Frankfurt) while
we worked heavily on the Cyberstorm PPC to finish them and had
no time to search for the Haage problem at once.
That愀 also what I told them personally on the phone when
they phoned me with more detailed informations about the problem.
4) That I was *never* impressed by the StormC quality is no
secret and Haage probably assumed that I惴 the only one which is
opposing against their control over the whole PowerUP
kernel,tools&app market.
Obviously this was a major fault on his side to assume I惴 his
only problem and now he愀 in big trouble.
In simple words...He pokered and lost.
By the way...aren愒 you the one which is now working for Haage&Partner
doing their Java thing ?.
Showing your statement in a very interesting light expecially as I扉e
never had any personal contact with you you claim things you could
never know about me.
which does their Java
--
Ralph Schmidt,la...@popmail.owl.de(private),NextMail welcome
Phase5 ,la...@ufoo.phase5.de(work),NextMail welcome
Brett Saunders
> : Amiga developers to use the StormC compiler as "The Standard"
> : which IMHO is a VERY BAD THING INDEED(tm)
>
> : fall for it.
>
> And at almost $700 AUS for the Storm C compiler, I personally wont be using
> it.. esp when the GNU utilities are free and generally of better quality IMO.
My experience with Storm C is simple:
I downloaded the demo.
I spent lots of time splitting it onto 1.4 mb disks.
I went home all excited to try it on my A4000.
I cated the archive.
I decompressed the archive.
I installed Storm C.
I ran storm C.
My system crashed.
I tried again.
and again.
and again.
Then I figured, oh yeah, it may be another program causing problems.
So I booted 3.1 in a bare minimal configuration.
I ran Storm C.
It crashed.
Do you think I would pay money for Storm C?
Do you think I would develop applications in Storm C?
Do you think I would pay money for WarpOS?
-Brett
Mans Engman
On 01 Oct 1997 14:39:03, Steffen Haeuser <mag...@birdland.es.bawue.de> wrote:
>
> ÿ° mo...@bulldozer.rydnet.lysator.liu.se schrieb : ÿ´
>
>Hi!
>
>mo> No, no, no. Running WarpOS means that P5 software, or software created
>mo> with P5's _free_ devkit won't run anymore. Such a solution is a big step
>mo> backwards, back to the good old days when you had to reset the computer
>mo> between running programs.
>
>Only, if you use the WarpOS V8+. WarpOS V7 (a wrap-around on ppc.library) will
>always be available, and it runs parallel with the Phase 5 things.
powerpc.library V7 -- Does NOT use WarpOS
powerpc.library V8+ -- Uses WarpOS
So WarpOS is the hardware-bashing kernel which sits in warp.library,
the different version numbers of powerpc.library is another story.
Or maybe I'm just confused.
Steffen Haeuser
ĸ° f94...@efd.lth.se schrieb : ĸī
Hi!
f9> If the development tools encourage peoble to write bad and slow code, more
f9> developers will do that. Just take Visual Basic as an example. If Visual C
I do not think that can be applied on WarpOS.
Steffen Haeuser
Steffen Haeuser
ĸ° w...@gf.phase5.de schrieb : ĸī
Hi!
wd> Do you think that the fact that we supplied free developer boards and
wd> all kind of support to them indicates that we "declined all
wd> cooperation"? They did not offer support, but frequently asked us to use
wd> their binary format and make Storm C the only and official development
wd> tool.
I only told what they were stating.
wd> In fact, they did develop their software and their strategy over a long
wd> time, and promised all the time that their solution would be compatible
wd> with the PowrUp system software. On the 25th of September they for the
wd> first time informed us about their WarpUp stuff, and started a massive
wd> marketing campaign the next day.
In fact with WarpOS V7 it is compatible.
>> ... i think the HAL to be used with PowerUP should be:
>> - efficient
>> - comfortable in usage
>> - support modern compilers
wd> That's what our HAL does. It even does support compilers from different
wd> vendors and on different platforms.
As to my information it is not the case (different platforms are not the
question... i do not want to buy a Unix system for developpement on my Amiga).
wd> I am sorry that the situation forces me to speak very frankly, but I
wd> don't see reasons to hide the truth in nice words.
But is there a need to do it in such a subjective (and partially maybe
incorrect) way like your first statement? Would a more neutral thing, based on
technical stuff, not have been better ?
wd> You should read the WarpUp announcement as well as the Amiga guides
wd> enclosed with WarpUp again, and reconsider if you believe the
wd> indications and frank comments about the phase 5 software solution are
wd> fair.
Fairer then the "reply", for sure.
wd> THEY say it is more efficient. THEY claim it to be faster or better. We
wd> deny this. And as it is similar in functionality, why should it take
Prove it !!! From H&P i got exact numbers, also from H&P there is a demo both
working on Native WarpOS and on a ppc.library Wraparound. From you ? "It is
wrong." As to the functionality, appearently it is much less fixed to the
current implementation of Exec, while ppc.library seems internally very fixed
to this. This might be a major problem, when Exec is no longer available or
changed.
Well, i would appreciate at best certain benchmarks, that would run on both
systems :)
Steffen Haeuser
Steffen Haeuser
ĸ° w...@gf.phase5.de schrieb : ĸī
Hi!
>> "Try our system. You'll see that it is better" is a valid argument.
>> "Use our system. You are not allowed to use another one" is none.
wd> As I outlined above, the second statement is not our position.
Does this (and the comment about being for free choice in a older mail) also
apply for WarpOS ? :)
Steffen Haeuser
James McArthur
Wolf Dietrich (w...@gf.phase5.de) wrote:
[ .. ]
: So, to answer your question: Yes. Everything that supports PowerUp is
: appreciated from our side, as long as it does not disable functionality
: that we provide and that our customers and developers want to use. We do
: also support developments such as Linux for the PowerUp, which is a
: completely different solution.
Well, I'm sold on the P5 approach, and I know what I'll be using to write all
of my code :)
Nils Henrik Lorentzen
Michael Rock wrote:
>
> Justin Smith wrote:
> >
> > Stefan Burstrom <f94...@efd.lth.se> wrote in article
> > <1276.7214...@efd.lth.se>...
> > > Conclusion: There will always be programs that disables the real PowerUp
> > > kernel just because some stupid developer though WarpOS was nifter at the
> > > expens of incompatible software.
>
> ... or that the deveoloper needs some low-level stuff that the powerup
> does not
> provide for, upto now.
>
> > I can understand why phase5 is pissed. They are delivering the PowerUp
> > boards with the promise that software developed for them (using the PowerUp
> > software) will run. Using WarpUp, H&P break this promise between phase5
> > and purchasers of PowerUp boards and software.
>
> The V7 uses their ppc.library, so it will not break any promise.
> But if you install the V12, that is your decision.
> So how can you say that the ones who did the software you installed is
> the
> one to break any promise ?
What you say here is that *all* apps compiled for whatever version
of WarpOS will work without disabling ppc.library, if the user
installs V7 and not a later version ?
There won't ever be apps compiled specifically for > V7 ? (eg. using
functionality that is only available WarpOS >V7)
If this is true, it totally changes the matter, since it
(as you mention) will be up to the user to decide wheher or not the
ppc.library should be disabled. And then I wonder: Why on earth
haven't someone mentioned this earlier ?
Anyways, I'm sure both systems has its upsides and downsides
compared to the other.
Wouldn't it then be cleverer to cooperate and create _one_ even better
product, instead of two (maybe incompatible) ones ? Both products
would AFAI have understood been given away for free, anyway,
so there is very little to loose economically by some cooperation.
Some obvious demands for such a system is that it
1) Can be made to work with *any* compiler (not just f.ex StormC).
2) Should not rely on *any* propriarity standard. AFAI have understood
neither the ELF format, nor the extended Amiga-hunk-format would be
propriarity. (By "propriarity" I mean that one has to pay royalties to
use it).
3) It should of course be made as effiecent as possible.
>From what I have seen, there seems to be two different "philosohys" here
(correct me if I have misunderstood something):
- Separating the PPC code tasks.
for: little CPU time is wasted in switching between the CPUs
against: programmers
- Doing a "quick and dirty" conversion of a program.
for: We can get a lot of powerup applications *at once*.
When/if a new PPC AmigaOS or clone appears, the PPC-compiled app
may be run on it without recompiling it again. (AFAI have
understood).
against: Wastes CPU time at each OS call, because of CPU switching,
and AFAI have understood, one need StormC to
recompile.
IMHO both solutions should be implemented in such a system, if possible.
IMO It could be OK that apps could in the second approacch
only be compiled with StormC, *if* the first approach also can be done
with
other compilers, like gcc.
Well, that was my contribution for getting the debate into a more
constructive direction.
I kinda get frustrated when I see more time spent on
bashing/spreading "facts" than on finding constructive solutions
to the problem. (And as I see it neither P5 nor H&P is any better than
the other)
>
> Michael Rock, student of computer science, TU Braunschweig.
>
> Also at
> Haage & Partner PPC-RISC Compiler department.
>
Nils Henrik Lorentzen
AROS participator <http://aros.fh-konstanz.de/aros/>
Martin Skowronski
Hello Michael Rock, on 02-Oct-97 16:41:13, you said,
>Hello #?.
>Per Jonsson wrote:
>> But teher is no response or reaction from H&P, if I got a better
>> product and got a lot of "shit" in a newsgroup I would scream.
>> I want to see some response from some at the H&P base.
>Statements are currently being written and translated.
>Nobody here expected this kind of over-reaction from P5.
I hope these "statements" are based on facts and not a knee jerk
reaction to the inflammatory comments made by Wolf Dietrich.
>
>> > What a waste of time, energy and probably good programmers, they should
>> > have done something positive for the Amiga.
>To speak for me, there was no waste of energy and time.
>And i have the strong feeling we did something positive for the
>amiga, but the whole story is a bit to complex to be written down here.
>
>> It's a pity. Instead of making WarpOS they could have joined P5 and
>> maybe optimized the CyberGL just a little bit more.
>What do you mean by "join" ?
>Ask for employment ?
Funny, but I think Wolf D's analogy of Porsche testing by driving it
backwards was funnier ;-)
>
>> I still hope this just is a big misunderstanding. I think it would be
>> the best thing.
>It looks like a combination of misunderstanding combined with
>over-reaction to me. But of course, i can only speak for myself.
This entire situation *should* never had happened. Without pointing fingers
this problem should have been addressed in private by the parties concerned.
Having failed to reach an agreement, the best thing would have been to have
phase5 and H&P release a joint statement outlining their differences in
a calm, rational and professional way. That's what I advise my clients when
they have disputes with partners engaged in joint ventures.
Regards,
Martin Skowronski
ProActive Images
www.proactive.net
PGP public key on request