Warpsness problems :(
Tom Miles
I've recently bought a blizzardpc card, and to celebrate I've been trying
out as many ppc programs I can get my hands on. Amongst those I've tried
is warpsnes, which I've so far been unable to get running.
I've used snoopdos to see what is going on, and I've discovered that the
program gets caught in an eternal loop opening libraries until my machine
runs out of memory. The exact sequence is:
ppc.library
intuition.library
utility.library
68060.library
libs:68060.library
libs:libs/68060.library
This repeats ad nauseum, with the libraries opening successfully each time.
My setup is:
603e+/040 A1200 OS3.0 CGFXAGA
warpsnes version 3.3
ppc.library version 46.19
powerpc.library version 14.1
68060.library 43.1
68040.library 44.4
I've tried emailing the author, but the address in the docs doesn't seem to
work. If anyone has any suggestions as to what may be going wrong then please
let me know.
TIA,
Tom
--
__
Http://www.teknogrebo.freeserve.co.uk : /// Amiga 1200T/040 25Mhz
UIN: 15835139 : __ /// 603e+ 233Mhz 18Mb RAM
Techno Grebo Skater Geek : \\\/// 4620Mb HDs 24x CDROM
Tatooed Vegan Piercing Freak : \XX/ Microvitec 17"
Steffen Haeuser
t...@teknogrebo.freeserve.co.uk wrote :
Hi!
to> Hi there,
to> I've recently bought a blizzardpc card, and to celebrate I've been trying
to> out as many ppc programs I can get my hands on. Amongst those I've tried
to> is warpsnes, which I've so far been unable to get running.
to> I've used snoopdos to see what is going on, and I've discovered that the
to> program gets caught in an eternal loop opening libraries until my machine
to> runs out of memory. The exact sequence is:
to> ppc.library
to> intuition.library
to> utility.library
to> 68060.library
to> libs:68060.library
to> libs:libs/68060.library
to> This repeats ad nauseum, with the libraries opening successfully each time.
to> My setup is:
to> 603e+/040 A1200 OS3.0 CGFXAGA
to> warpsnes version 3.3
to> ppc.library version 46.19
to> powerpc.library version 14.1
to> 68060.library 43.1
to> 68040.library 44.4
to> I've tried emailing the author, but the address in the docs doesn't seem to
to> work. If anyone has any suggestions as to what may be going wrong then please
to> let me know.
The address works fine. But i never received an email by you.
Well, i suggest that it is a termination problem (that infamous "Anti-WarpUP"
Code by Phase 5)
Set env:powerpc/TERMINATOR to "2" and don't start any ppc.library apps during
bootup (this also applies to CGXAGA... the only difference between CGX 68k and
PPC is that PPC disables WarpUP... okay, for AGA no CGX 68k exists... but for
GFX Boards this is like this...).
Also go sure that you add enough stack (300000 are enough FOR SURE :) )
WarpSNES does not open ppc.library AT ALL, BTW...
BTW: Which versions have your warp.library and warpHW.library ?
Steffen
Gerrit-kjeld Dusseljee
On 19 Dec 1998 04:35:35 Steffen Haeuser wrote about "Warpsness problems :(":
Well here goes 'mr.ButtKiss H&P' again. Did you recieve more money for your
PR-campaign?
>
> Set env:powerpc/TERMINATOR to "2" and don't start any ppc.library apps during
> bootup (this also applies to CGXAGA... the only difference between CGX 68k
and
> PPC is that PPC disables WarpUP... okay, for AGA no CGX 68k exists... but for
> GFX Boards this is like this...).
It's the unstable nature of a 'reverse ingeneered not approved M$ way' WarpUP.
>
> Also go sure that you add enough stack (300000 are enough FOR SURE :) )
>
> WarpSNES does not open ppc.library AT ALL, BTW...
>
> BTW: Which versions have your warp.library and warpHW.library ?
>
> Steffen
--
A4000/060@50/604e@200/118MB ram/4.3GB hd/12xCDrom/Toccata/Hypercom3Z
--Member of:Convergence International- The NON-Wintel user group--
--Cracking RC5-64// Warp*S - a program of M$ Tactics Inc.
ICQ:255060020
E...@thepentagon.com
Gerrit-kjeld Dusseljee <dus...@xs4all.nl> sayeth:
>
> Well here goes 'mr.ButtKiss H&P' again. Did you recieve more money for your
> PR-campaign?
Good roar, lion; just not LOUD enough (;
Watch out for the 100-line-essay he`ll probably send to your eMail account.
Maybe we should post a weekly "Steffen H. FAQ" too ?
> It's the unstable nature of a 'reverse ingeneered not approved M$ way' WarpUP.
Burn baby, BURN !
Greetings to all WarpOS-Fetishists ... NOT !
ek
-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own
Steffen Haeuser
E...@ThePentagon.com wrote :
Hi!
EK> In article <75flar$9ib$4...@news2.xs4all.nl>,
EK> Gerrit-kjeld Dusseljee <dus...@xs4all.nl> sayeth:
>>
>> Well here goes 'mr.ButtKiss H&P' again. Did you recieve more money for your
>> PR-campaign?
EK> Good roar, lion; just not LOUD enough (;
Maybe too loud. Actually i have only the future of the Amiga in mind. Which
appearently you guys don't care for.
Actually i am too big an Amiga-Fan, as that you can hurt me with your insults.
For the Amiga i am even taking this.
And that has nothing to do with H&P... if Phase 5 would be the ones doing
WarpUP i would be pro-Phase 5 :) It has only todo with the PRODUCT, which is
the best for the Amiga.
To come to something else... why are you two talking so much pro ppc.library,
while it is obviously doing harm to the future of the Amiga ? Are you getting
payed by Phase 5 ? Shame on you...
>> It's the unstable nature of a 'reverse ingeneered not approved M$ way' WarpUP.
EK> Burn baby, BURN !
You do not actually BELIEVE this crap ?
Why would then WarpUP have kernel-features inside always BEFORE ppc.library
have them ? Sounds more like it would be the other way round, wouldn't it ?
That is probably not the case, though, as Ralph Schmidt would have problems
copying stuff from WarpUP, as it uses PowerOpen while his system uses Server
V... they can't be just like that copied.
Also they use totally different approaches. ppc.library "hack's" into the
exec Kernel (and spawns then off a PPC Native Server Process, after it was
hacked in) while WarpUP runs from the start as totally independent PPC Kernel,
having nothing to do with exec.
Steffen Haeuser
Ed Collins
> > It's the unstable nature of a 'reverse ingeneered not approved M$ way' WarpUP.
>
> Burn baby, BURN !
>
> Greetings to all WarpOS-Fetishists ... NOT !
Get a life you two.
Ed.
--
________ _________
\___ \ / \ / ___// HOME http://www.d-n-a.net/users/dnetEFoE/
| \ \\_/\_// __// CEG: http://www.explorer2260.home.ml.org/
|_ /__\____/_// TWF: http://www.worldfoundry.home.ml.org/
The World Foundry - Amiga PPC - Explorer 2260
E...@thepentagon.com
"Ed Collins" <coll...@dnet.co.uk> wrote:
> > > It's the unstable nature of a 'reverse ingeneered not approved M$ way' WarpUP.
> >
> > Burn baby, BURN !
> >
> > Greetings to all WarpOS-Fetishists ... NOT !
>
> Get a life you two.
>
> Ed.
> --
I already have one. It`s backed up on CD-R somewhere (I don`t need it
everyday).
<sad>
There are other people who most urgently need a life.
</sad>
Alan L.M. Buxey
: Well here goes 'mr.ButtKiss H&P' again. Did you recieve more money for your
: PR-campaign?
?? why dont you s.t.f.u. before posting such ignorant remarks. It is
well known that in the 46.19 ROM, there are extra lines of code to stunt
the usage of WarpUP (apparently done before the P5/H&P 'agreement') -
there was a workaround created by H&P to circumvent this...I'm not
worried that Steffen makes this public (although so often might seem a
bit evangelistic)
: It's the unstable nature of a 'reverse ingeneered not approved M$ way' WarpUP.
?? what's M$ got to do with this? H&P did not need to reverse engineer
- the logic required for the PPC cards is quite elemental if you know
your stuff (ie PPC's, logic etc) WarpUP is very stable on my machine.
alan
Alan L.M. Buxey
: Well, i suggest that it is a termination problem (that infamous "Anti-WarpUP"
: Code by Phase 5)
talking of which - the new BVision card requires a small update
(FlashROM) to 46.21...are there any more strange changes during this
update that you have discovered?
: WarpSNES does not open ppc.library AT ALL, BTW...
: BTW: Which versions have your warp.library and warpHW.library ?
ah! maybe he's using v7 of the warpUP system
alan
Alan L.M. Buxey
: > Get a life you two.
: I already have one. It`s backed up on CD-R somewhere (I don`t need it
: everyday).
mines a better life then - I had to back it up onto DVD instead :-)
(PS On my Amiga using latest IdeFix97 and BLAZEMONGER-DVD.DEVICE
which allows instant parallel access to all of the DVD and fullscreen
HDTV playback audio+video in 32bit even on OCS!!!)
alan
Nathan Wain
On 21 Dec 1998, Alan L.M. Buxey wrote:
> mines a better life then - I had to back it up onto DVD instead :-)
:-) Well, he just gave me the impression that his life is a waste
of space, and nothing more.
> (PS On my Amiga using latest IdeFix97 and BLAZEMONGER-DVD.DEVICE
> which allows instant parallel access to all of the DVD and fullscreen
> HDTV playback audio+video in 32bit even on OCS!!!)
Really? What have you used it for? And how well did it work?
Details! Details!!!
Nathan.
Steffen Haeuser
kc...@central.susx.ac.uk wrote :
Hi!
kc> : Well, i suggest that it is a termination problem (that infamous "Anti-WarpUP"
kc> : Code by Phase 5)
kc> talking of which - the new BVision card requires a small update
kc> (FlashROM) to 46.21...are there any more strange changes during this
kc> update that you have discovered?
Hmmm, i would be interested in infos about this too !!! (Most likely, though,
it is just the same patch like the one WarpUP needed from 3.0->3.1 ... adding
support to access the VideoRAM of the Non-Standard GFX Board... as i remember
people owning CS-PPC needed such a patch too for CV/PPC...)
kc> : WarpSNES does not open ppc.library AT ALL, BTW...
kc> : BTW: Which versions have your warp.library and warpHW.library ?
kc> ah! maybe he's using v7 of the warpUP system
That would explain it. And i told before: I do not guarantee that ANY of my
software will run below V14.
V7 is only sitting on top of ppc.library, and can because of this only
implement features which are present in ppc.library. A lot of features
concerning task-stuff need powerpc.library V9/10/11 or 14 at least... But then
V7 was done under the impression, that Phase 5 would "adapt" their code to
some extent, if WarpUP would enable a ppc.lib-compatible solution (which of
course did NOT happen). Well, luckily ppc.library is now one thing more for
the "OUT" list in your mail before (or had you that one in mind when
mentioning "proprietary stuff" ? :)
Anyways, i tested WarpSNES on a lot of PPC-Systems, including A1200-PPC, and
never had problems with it. As i told before, i already had a LOT of problems
with ppc.library software (on some of the systems it ran... but not on all...).
Steffen Haeuser
Steffen Haeuser
kc...@central.susx.ac.uk wrote :
Hi!
kc> ?? why dont you s.t.f.u. before posting such ignorant remarks. It is
kc> well known that in the 46.19 ROM, there are extra lines of code to stunt
kc> the usage of WarpUP (apparently done before the P5/H&P 'agreement') -
kc> there was a workaround created by H&P to circumvent this...I'm not
kc> worried that Steffen makes this public (although so often might seem a
kc> bit evangelistic)
Okay, right :) But i really simply have enough of that discussion.
But the problem is: I can leave standing a clever remark, but when i read
something extremely silly, like what the guys whom you corrected in your above
mail are writing i can't help, but MUST reply. Sadly, some people with certain
"half-knowledge" are writing the whole nonsense all over again - and are in
fact fighting the people who have the best for the Amiga in mind. And i am
convinced they did not mean it like that...
I definitely do not want that silly discussion where everyone has enough of it
AGAIN. But well, maybe we should fix it once and for all :)
kc> : It's the unstable nature of a 'reverse ingeneered not approved M$ way' WarpUP.
kc> ?? what's M$ got to do with this? H&P did not need to reverse engineer
kc> - the logic required for the PPC cards is quite elemental if you know
kc> your stuff (ie PPC's, logic etc) WarpUP is very stable on my machine.
Okay, io feel that as an invitation to add the things together. Actually, you
COULD bring M$-like behaviour into talks, but NOT together with the name H&P.
With the name P5 this maybe could be - right - to bring M$-like behaviour into
context. Let's check the fact.
1. H&P told P5 at the start that they would prefer a more Amiga-like
integration of the PPC, and that they would do this for P5 if they liked (P5
replied "Do it our way or do it no way".
2. (Well, maybe this does not belong here... but at about the same way Sam
Jordan contacted P5 and told he would be interested in programming the PPC.
Based on PowerOpen - which today WarpUP - is using - he had started
developpement of a PPC Assembler. He never got a reply from P5... H&P on the
other hand replied to his mail at once)
3. Independent people (like me at the time) where told definitely wrong things
by P5 like "the powerpc.library" - like WarpUP was still called at this time -
will not run on the release version Boards.
4. Phase 5 started offering developpers free PPC Boards, if they signed a
contract where they had to tell they would "never support non-Phase5-compliant
Software". WarpUP and Picasso96 were called by name as such software. Of course
it was preffered that these treaties were kept private. It is interesting:
Microsoft at a time used similar treaties !!!
5. Then, when WarpUP was released, they wrote some garbage-text about it,
which was really ABSOLUTELY wrong. They claimed WarpUP being slower, being
based on reengineering (that this is not the case can be tested easily...
WarpUP uses PowerOpen, not Server V, so a reengineered ppc.library would be of
no use... on the other side a lot of features appeared on WarpUP *before* they
appeared in ppc.library. Also i know that Sam Jordan based his work on three
things:
(1) The Documentation of Motorola of the PowerPC
(2) Basic Operating System knowledge
(3) Some register-numbers (these were got by reengineering, right, but this
is absolutely no critical stuff... These are 5 numbers or such... where
the register-addresses of the PPC are mapped to memory... on the other
hand he did a much better job concerning the Kernel)
The real work about a new Kernel is - the KERNEL. And when you compare the
featurelist of both kernels you will notice who supports "more".
6. Later, when the A1200 PPC was released, Phase 5 used a 256 KB ROM instead
of a 128 KB ROM - raising the price for the end user in the progress !!! -
just to disable WarpUP. They also announced a clearly vaporware system,
just to try to move developpers away from WarpUP.
7. Later when the M&M Team decided to do their Game for 68k and WarpUP, they
contacted the publisher of the team and tried him to force the team into
ppc.library. The result was the cancellation of M&M (okay, the team managed to
convince their publisher for WarpUP, but then came the SECOND
WarpUP-Disable-ROM which was released, and as a consequence of THAT the M&M
team left the Amiga, where such inprofessional things happened...). This team
was not the only one were such a thing threatened to happen, but in fact the
only team where the team gave up fighting the M$-like practices.
8. At the WOA P5 promised to "remove the Anti-WarpUP-Code of the ROM with the
next release". H&P in return promised that they would do a ppc.library
Emulation into the next version of WarpUP (provided Phase 5 provided the
needed technical data). What happened: The next version of the FlashROM
disabled WarpUP again (this time by modifying the LoadSeg in a way that it
crashed, if ppc.library could not be opened), and they declined to give out
any technical data about ppc.library. As a consequence of these things, H&P
cancelled the ppc.library Emulation.
Well, this is the story.
And please, let this now be an end to this thread.
And to clarify one more thing once and for all: I did not take any money of
H&P (nor did they offer me any). At the start i examined the two PPC Kernels,
and while being at the start more pro ppc.library - because P5 people told me
powerpc.library would not run any longer with the release version PPC Boards,
the typical sort of P5-statements, and there i still believed them - i soon
was definitely pro-WarpUP. And this did not change then. Neither am i in ANY
way working for H&P. I am working for Quark, not H&P... my Amiga-Stuff is done
in my freetime.
Steffen Haeuser
E...@thepentagon.com
kc...@central.susx.ac.uk (Alan L.M. Buxey) wrote:
>
> : I already have one. It`s backed up on CD-R somewhere (I don`t need it
> : everyday).
>
Well, I just forgot to tell you how much CD-R`s a complete backup takes (;
(Hint: I need a whole disk for the short-term memory only.)
So, does size matter ??? DISKSPACE, not what you think ! (;
If so, I can`t understand all the numerous "Amiga vs. PeeCee" discussions,
because then Widows 59/89 of course would be far superior to AmigaOS,
just because it uses *much* more diskspace.
Can anyone tell me how much space a complete Widows installation eats up ?
Or should I ask: How much will be left from a 4 gig drive ???
I`m already wondering if Widows 2000 B.C. will be DVD only ...
> (PS On my Amiga using latest IdeFix97 and BLAZEMONGER-DVD.DEVICE
> which allows instant parallel access to all of the DVD and fullscreen
> HDTV playback audio+video in 32bit even on OCS!!!)
>
w0w !
> alan
Oh, btw, I`ve never (well, almost) criticized cB,
and I`ve never (NEVER !!!!!!) criticized p5.
(Far from it: I am a satisfied user of both cB soft and p5 hard ...)
Thomas Johnsrud
>installer didn't work properly, and I didn't have warphw.library installed.
>I managed to get the warpsnes up and running to the point of loading in a
>rom, but I didn't have enough memory to get a game running on it. It seems
>my 16 meg just doesn't cut the mustard when it comes to ppc programs :(
>And I missed the cheap RAM of last month :((
Go to: http://come.to/snes9x/
And get the PowerUP version that is much better and newer.
--
Cool T.J/Sapphire - flek...@online.no - A1200/040/PowerPC + A4000/040
Steffen Haeuser
flek...@online.no wrote :
Hi!
>>rom, but I didn't have enough memory to get a game running on it. It seems
>>my 16 meg just doesn't cut the mustard when it comes to ppc programs :(
>>And I missed the cheap RAM of last month :((
fl> Go to: http://come.to/snes9x/
fl> And get the PowerUP version that is much better and newer.
We will talk again when WarpUP runs on the Escena-Board, while ppc.library
version still on the PowerUP one :)
And yes, i will adapt the latest SNES9x sources into it later... currently i
am busy with a version of ADescent which supports BOTH WarpUP and Warp3D !!!
BTW: The only case where ppc.library version of SNES is better is on AGA
crap... on GFX Board Systems my version is faster...
Steffen Haeuser
Steffen Haeuser
flek...@online.no wrote :
Hi!
fl> Go to: http://come.to/snes9x/
fl> And get the PowerUP version that is much better and newer.
And also i just checked, and it is not true !!! There is no later version...
And besides... what about quitting ppc.library and coming back to *the Amiga*
???
Steffen Haeuser
Thomas Johnsrud
>Hi!
fl>> Go to: http://come.to/snes9x/
fl>> And get the PowerUP version that is much better and newer.
>And also i just checked, and it is not true !!! There is no later version...
Jupp, cause it is based on the version 1.10 of the Snes9x. That is *not* true
with the WarpSnes.
>And besides... what about quitting ppc.library and coming back to *the Amiga*
>???
?
I run the ppc.library all the time, there is no problem.
Thomas Johnsrud
>>>rom, but I didn't have enough memory to get a game running on it. It seems
>>>my 16 meg just doesn't cut the mustard when it comes to ppc programs :(
>>>And I missed the cheap RAM of last month :((
fl>> Go to: http://come.to/snes9x/
fl>> And get the PowerUP version that is much better and newer.
>We will talk again when WarpUP runs on the Escena-Board, while ppc.library
>version still on the PowerUP one :)
>And yes, i will adapt the latest SNES9x sources into it later... currently i
>am busy with a version of ADescent which supports BOTH WarpUP and Warp3D !!!
>BTW: The only case where ppc.library version of SNES is better is on AGA
>crap... on GFX Board Systems my version is faster...
Bah! On my machine the WarpUp version, always crashes!
Steffen Haeuser
flek...@online.no wrote :
Hi!
>>BTW: The only case where ppc.library version of SNES is better is on AGA
>>crap... on GFX Board Systems my version is faster...
fl> Bah! On my machine the WarpUp version, always crashes!
Then this is a problem of your machine, i would say :)
Well, without detailed bug-report two things come to my mind:
1) Termination-Problem
ppc.library patches (and on A1200 PPC it is always active) the LoadSeg
function which is used for Library-Loading. WarpSNES uses a PPC Shared
Library, which disables (using the env:powerpc/TERMINATOR setting)
ppc.library. So the LoadSeg-Patch of Phase 5 just crashes, when loading
rtgmaster. This is not the fault of WarpSNES or rtgmaster, but of the
Load-Seg-Patch (they should not assume that ppc.lib can always be opened). To
fix it:
setenv powerpc/TERMINATOR 2
(for newer FlashROMs, at least, for old ones it was 1)
Do not install any ppc.library software enabled during bootup, including CGX
PPC, ppc.library based Datatypes, PPCInstall, ElfLoadSeg, RC5-PPC (else the
LoadSeg-Patch cannot be disabled)
Use a stack of at least 30 KB.
If you have MCP installed (rtgmaster 68k part, the initialization, has a bit
problems with it) you should install MCPramlibpatch, by putting
MCPramlibpatch >nil:
into startup-sequence.
If you own a BV/PPC, do:
setenv powerpc/force 1
setenv powerpc/gfxaddr $e0000000
to enable direct Video RAM Access to the Non-Standard Board
Basically the problem above was because of P5 wanting to choose for YOU which
Kernel you HAVE to use. They even used a bigger (more expansive) FlashROM
because of that, that you cannot choose freely yourselves. Luckily WarpUP
found a way around this...
Steffen Haeuser
Steffen Haeuser
flek...@online.no wrote :
Hi!
fl> Jupp, cause it is based on the version 1.10 of the Snes9x. That is *not* true
fl> with the WarpSnes.
WarpSNES is also based on 1.10. At least the latest version.
>>And besides... what about quitting ppc.library and coming back to *the Amiga*
>>???
fl> ?
fl> I run the ppc.library all the time, there is no problem.
Still it has not much to do anymore with the Amiga :)
(And crashes for many people's systems...)
Steffen Haeuser
Jyrki O Saarinen
> Basically the problem above was because of P5 wanting to choose for YOU
> which Kernel you HAVE to use. They even used a bigger (more expansive)
> FlashROM because of that, that you cannot choose freely yourselves.
> Luckily WarpUP found a way around this...
Yes, it is always easier to blame others.
Sure, there are no bugs in rtgmaster, which crashed for me on almost
every app using it (demos) or they didn't work. I don't see a reason
for using such a library either, it doesn't offer anything which
any Amiga programmer couldn't do in few minutes.
Oh, if the ppc.library in BPPC flash is such a problem, why the
hell Linux/APUS is able to work on BPPC; it manages to get rid
of it?
Sam Jordan
> Oh, if the ppc.library in BPPC flash is such a problem, why the
> hell Linux/APUS is able to work on BPPC; it manages to get rid
> of it?
Because Linux doesn't require AMIGA-OS to work in parallel! But
WarpOS does (because the PPC-only mode only works in blackbox
emulation mode) and in dual-CPU mode it is necessary to have
a 68K part running for communication.
The problem with WarpUp and P5-Flash-ROM is not, that WarpUp
doesn't work, but that AMIGA-OS doesn't work anymore if the
ppc.library is not in the system's library list. The *AMIGA-OS*
function dos/LoadSeg crashes, if WarpUp attempts to start and
to remove the ppc.library from the system's library list (as
you know, both kernels can't run together, so one must not
be active). The reason for the crash is, that the patch simply
doesn't check the return value of OpenLibrary("ppc.library",version).
I really wonder, why everyone always reports WarpUp crashing
by pointing to the 68K failure requester. WarpOS is a *PPC*
kernel, and the 68K part only serves for communication. The 68K-
Part is *NOT* = WarpOS. It is meanwhile proven that WarpOS can
run completely independently from the 68K (in blackbox
emulation mode) so all problems which happen on the 68K
side, are not WarpOS problems.
Most problems even happen, *before* the PPC kernel even has
started, and nevertheless people are complaining about WarpOS
beeing buggy. WarpUp has problems to start, because there are
so many obstacles to bypass. As soon as the startup problem is
solved, the kernel itself is running very very reliable.
Comparing the ppc.library and WarpOS reliability when the
author of the first system has all possibilities to make
the life of the other system's author as difficult as possible
(either by intention or not) is at least doubtful. And this
flash ROM did make life heavily difficult for us, since it
happened already two times, that the flash ROM did disable
all WarpUp applications. If the flash ROM wouldn't contain
the ppc.library, *MOST* problems would disappear *IMMEDIATELY*.
(in fact this possibility is either or will be possible soon,
by skipping the ppc.library after warm reboot, such a tool
is or will be available).
No I'm asking you why everyone is hacking us because of
the WarpUp problems, when we are not responsible for these!
It's simply a dependency between ppc.library and AMIGA-OS
which is causing these heavy problems. And why the hell
didn't they simply adapt the CSPPC solution (ppc.library
in LIBS:) for the BPPC? Maybe the well-known text at the
end of the BPPC flash ROM might answer this question. Most
CSPPC users don't have problems with WarpUp, except if they
did something wrong and didn't want to mail me to ask for
solutions (I'm usually responding immediately after mails
come in).
Again: WarpUp is *running* very stable. It has problems
with *starting*. Please never confuse this.
Please please stop this fake senseless stupid dummy 'why does
Linux work and WarpUp doesn't' argument (there is no dos/LoadSeg
in Linux...). In this special case you definitely don't know what
you are talking about (I do never generalize the others competency,
because everyone is competent in different topics, but now I'm
referring to this specific case, where this is definitely not
the case).
bye
--
Sam Jordan
Steffen Haeuser
jxsa...@cs.Helsinki.FI wrote :
Hi!
jx> Steffen Haeuser <mag...@birdland.es.bawue.de> wrote:
>> Basically the problem above was because of P5 wanting to choose for YOU
>> which Kernel you HAVE to use. They even used a bigger (more expansive)
>> FlashROM because of that, that you cannot choose freely yourselves.
>> Luckily WarpUP found a way around this...
jx> Yes, it is always easier to blame others.
Well, but there IS Anti-WarpUP-Code in that FlashROM, that is a FACT. And have
a look at the ROM-code... what did change ? They put ppc.library and 680x0.lib
into ROM...
jx> Sure, there are no bugs in rtgmaster, which crashed for me on almost
jx> every app using it (demos) or they didn't work. I don't see a reason
jx> for using such a library either, it doesn't offer anything which
jx> any Amiga programmer couldn't do in few minutes.
Well, at the time it was great. Especially games coders which did not own GFX
Boards were always reluctant to support GFX Boards. A lib which enabled them
to test the GFX Board Code on their AGA Systems helped there much. (I know for
sure, that for example NAPALM would be a AGA-only game if i had not convinced
the author for a rtgmaster version.... the final argument in this was "Even if
it is GFX Board Code you can run it on your AGA system, so no problems with
testing".
jx> Oh, if the ppc.library in BPPC flash is such a problem, why the
jx> hell Linux/APUS is able to work on BPPC; it manages to get rid
jx> of it?
Well, WarpUP manages this too. But some people seem to complain that they have
to set ENV-Variables. Also the situation is a bit different. Linux does not
run parallel to AmigaOS anyways, so nobody TRIES running Linux and Amiga
applications at the same time. For WarpUP and ppc.library this is tried, and
as soon as it is trieds, the problems appear (ppc.lib patched LoadSeg, as soon
as a ppc.lib program is started. The patched version then runs a ppc.library
function (which causes a crash of course if WarpUP shut down the ppc.library).
To fix this problem the termination is used, but termination only works if no
ppc.lib apps are currently running.
BTW: These problems ONLY appeared with the BlizzardPPC (as the CyberstormPPC
does not have the ppc.lib in ROM) and can be fixed by setting one ENV-Var. and
throwing out ppc.lib apps of the startup.
The second problem is CGX PPC. If it is installed (at least the version i once
tried) and a WarpUP-Program opens a screen, it crashes during OpenScreen.
(Promotion is already running, so probably another patch... with CGX 68k or
P96 everything is fine). Ah, CGX PPC, noticed something else... a friend of
mine wanted to move from CGX PPC to P96. But the Picasso96Mode program did not
open the window !!!! Well, after we renamed cgxsystem.library and disabled CGX
PPC it came up fine... strange...
Of course possible that this was an old version and such things don't happen
anymore with recent versions, he had a very old version installed (it did not
run stable also, which was why he mainly wanted to switcvh to P96...)
Ah, and it does not help you to try to discuss all this away. At least the
"Anti-WarpUP" protection exists. (What happened with this CGX-thing i didn't
go deeper into...). But well, i am asking myselves why you are defending such
a firm. Also, if we think of that ppc.lib has not much to with Amiga anymore
(from big OS design things till down to big bloated executables, which never
were the way of the AmigaOS). You used to be a Amiga Fan a long time ago. I
find it sad that this changed. :(
Steffen
jx>
Rocou
>>Hi!
>>>>rom, but I didn't have enough memory to get a game running on it. It seems
>>>>my 16 meg just doesn't cut the mustard when it comes to ppc programs :(
>>>>And I missed the cheap RAM of last month :((
fl>>> Go to: http://come.to/snes9x/
fl>>> And get the PowerUP version that is much better and newer.
>>We will talk again when WarpUP runs on the Escena-Board, while ppc.library
>>version still on the PowerUP one :)
>>And yes, i will adapt the latest SNES9x sources into it later... currently i
>>am busy with a version of ADescent which supports BOTH WarpUP and Warp3D !!!
>>BTW: The only case where ppc.library version of SNES is better is on AGA
>>crap... on GFX Board Systems my version is faster...
TJ> Bah! On my machine the WarpUp version, always crashes!
On my machine too. Only H&P's demos work.
""msaupe@saale-net.de
> But well, i am asking myselves why you are defending such
> a firm.
Because some people see that without them we wouldn't have PPC at all!
And I highly doubt that another company would have done a PPC board;
don't forget that without the P5 boards there wouldn't be WarpUp
for the Escena board, yet.
>Also, if we think of that ppc.lib has not much to with Amiga anymore
>(from big OS design things till down to big bloated executables, which
>never were the way of the AmigaOS). You used to be a Amiga Fan a long
>time ago. I find it sad that this changed. :(
Even if WarpUp would be 1000 times better, you can't say ppc.library
is no Amiga. From a users POV I don't care which kernal is used
because both run stable (although of course it would be better if
there would have been only one kernal from the start).
WarpUp will very probably be in OS 3.5 - Okay, good.
Most new PPC programs will be for WarpUp - No problem
But the PowerUp programs that exist won't stop working (we can talk about
the Escena board when they arrive) and they are Amiga programs.
Bye
PS: If ppc.library is no Amiga then tell me for which System I bought
Reflections (ATM most professional program available for Amiga PPC I think)?
--
- Marco Saupe - Team *AMIGA* - http://www.saale-net.de/homepages/m.saupe -
------ msa...@saale-net.de --- Marco...@in.stud.tu-ilmenau.de -------
------ AMIGA 1200T PPC 603/200 -------- Amiga Zentrum Thueringen -------
Sam Jordan
> > But well, i am asking myselves why you are defending such
> > a firm.
>
> Because some people see that without them we wouldn't have PPC at all!
> And I highly doubt that another company would have done a PPC board;
> don't forget that without the P5 boards there wouldn't be WarpUp
> for the Escena board, yet.
Everyone knows what they have done for the AMIGA and that they did
important things. The problem is just, that many people think, that
this allows them to do what they want, it would give them some special
rights (and many people actually supported their Anti-WarpUp methods
*because* of their engagement in past, it was 'legimated'). We should
carefully separate the present from the past by not forgetting, what
they brought for good stuff for the AMIGA and by not tolerating that
they just misuse this image in present. It is always good to watch a
company carefully, because time changes, people change, philosophy
changes, everything changes. That's what many people don't want to see.
In case of Phase5 we have the situation that they actually based many
of their actions on their image as 'saver of the AMIGA', so they were
able to do things, which any other company never could do, since the
AMIGA users wouldn't tolerate that . But they of course tolerated the
P5 methods, because they were the 'saver of the AMIGA'. IMHO it is
dangerous to give a company such possibilities just because of
past engagement, because things could easily go out of control. Many
(not all) people had some kind of 'blind believing' (sorry for this
bad translation) by simply refusing to put their methods in question
("...a company which did all this good can't do mistakes...").
We should always put everything in question and reanalysing the
situation all the time, regardless of what was earlier. Just basing
everything on the past is dangerous, the chance of misusement gets
too high.
Therefore I don't see any legitimation to defend Phase5's Anti-Warpup
actions (with all these flash ROM problems, which easily could have
been solved by Phase5) just based on their image as 'saver of the
AMIGA'.
I also don't see any reason, why all this still continues. It would
be very easy just to drop the ppc.library from flash ROM and to put
it into LIBS:, since that's exactly the case on CSPPC boards. They
could do all AMIGA users a major favour, it would eliminate many
problems and the users still could use both systems, because they
have the freedom of choice. And it becomes more and more important,
especially since the upcoming W3D Permedia2 support requires a reliable
working WarpUp system.
bye
--
Sam Jordan
Steffen Haeuser
""msa...@saale-net.de wrote :
Hi!
>> But well, i am asking myselves why you are defending such
>> a firm.
""> Because some people see that without them we wouldn't have PPC at all!
""> And I highly doubt that another company would have done a PPC board;
And this allows them to destroy the market ? I don't think so. You also have
to think that most of the people who designed the PPC Board left the firm
or were fired in the last months...
BTW: I know of several firms who were interested in doing such a Board. But P5
had the market for PPC Hardware in their hands, so quite some PPC-projects
were cancelled.
Of course it was a great thing that they did the PPC Boards. But that we have
to accept ANYTHING from them in return, with this i do NOT agree. Neither i do
agree with all H&P says (i always was of the opinion that StormC was too
expansive... well, at least before StormC DSK release, with the expanded
functionality i find the price agreeable...)
""> Even if WarpUp would be 1000 times better, you can't say ppc.library
""> is no Amiga. From a users POV I don't care which kernal is used
I can say. Some things:
- I connect AmigaOS also very close with it's hunk format. It is a
resource-saving SMALL format, while ELF is big and bloated. Weren't we
always laughing about the PC-guys with their huge memory-needs and that ?
With ppc.lib it is their turn to laugh back :)
- Look at the API. For WarpUP really only some "PPC" was appended to the names,
asides from that it looks 1:1 like the original API. Not true for ppc.lib
- for most cases - it looks as if Ralph Schmidt invented stuff "ad hoc"
as he needed, without looking at the consequences.
- No possibility to integrate it to 68k-OS by use of a 68k Emulator
- Compare the implementation of PPC Libs with 68k Libs for both Kernels.
For WarpUP it is a 1:1 implementation - for ppc.library something completely
different !!!
""> WarpUp will very probably be in OS 3.5 - Okay, good.
""> Most new PPC programs will be for WarpUp - No problem
""> But the PowerUp programs that exist won't stop working (we can talk about
""> the Escena board when they arrive) and they are Amiga programs.
They will on Escena Board. At least most of them (as only around 10% run on
the ppc.lib Emulation)
""> PS: If ppc.library is no Amiga then tell me for which System I bought
""> Reflections (ATM most professional program available for Amiga PPC I think)?
For Laire-OS. :) (Well, this is a synonym used by some commercial
Amiga-Developpers, it means at the same time "ppc.lib stuff" and "bad coded
software", you have to get the rest of the sentence to find out which of them
was meant this time...) (Well, seriously... for ppc.library you bought it...)
Steffen Haeuser
Teemu Suikki
Steffen, grow up... You have been corrected before, and still you make
these same incorrect claims that you made a year ago..
I personally won't care which kernel is used, but it's annoying to see the
same crap over and over again.
>- I connect AmigaOS also very close with it's hunk format. It is a
> resource-saving SMALL format, while ELF is big and bloated. Weren't we
Duh. Hunk format is designed with 68k in mind, has known bugs and doesn't
support the split relocation needed for PPC code.
ELF format is universal, quite clean format, and supports anything needed
and much more. PPC programs don't run on 68k processor anyway, so why would
it be necessary to use a 68k file format?
ELF files are slighly bigger than hunk files, but "big and bloated" is
quite an over-estimation.. I compiled some test programs with scppc using
hunk and ELF formats, and there's only a few kilobytes difference.
>- Look at the API. For WarpUP really only some "PPC" was appended to the names,
> asides from that it looks 1:1 like the original API. Not true for ppc.lib
This shows that you _never_ have even take a look at the ppc.library API..
Here's a short list of ppc.library functions which only have "PPC"
appended:
PPCAddHead, PPCAddTail, PPCAllocMem, PPCAllocPooled, PPCAllocSignal,
PPCAllocVec, PPCAttemptSemaphore, PPCClose, PPCCloseLibrary,
PPCCoerceMethodA, PPCCreatePool, PPCCreatePort, PPCDeletePool, PPCDeletePort,
PPCDoMethodA, PPCDoSuperMethodA, PPCEnqueue, PPCFindName, PPCFindTagItem,
PPCFindTask, PPCFreeMem, PPCFreeSignal, PPCGetTagData, PPCInsert,
PPCNextTagItem, PPCObtainSemaphore, PPCOpen, PPCOpenLibrary, PPCRawDoFmt,
PPCRead, PPCReleaseSemaphore, PPCRemHead, PPCRemove, PPCRemTail, PPCSeek,
PPCSetSignal, PPCSignal, PPCWait, PPCWaitPort, PPCWrite
I only included the names which were exact matches and which I was sure of,
there are many "similar" functions like PPCGetMessage vs. GetMsg...
>- No possibility to integrate it to 68k-OS by use of a 68k Emulator
That's wrong. Both warpup and ppc.library implement 68k functions the same
way; you call a kernel function which then runs the 68k code. On current
systems this is done by the 68k processor.
The ppc.library API is quite clear here.. 68k emulation would only require
changes in PPCCallOS() and PPCCallM68k() kernel functions.
>- Compare the implementation of PPC Libs with 68k Libs for both Kernels.
> For WarpUP it is a 1:1 implementation - for ppc.library something completely
> different !!!
That depends which libs you are talking about. With SAS/C-PPC you can
create hunk-style libraries which can be called from both 68k and PPC code.
They are exactly the same format as regular hunk libs.
Then, you can have unix-style load-time binded libs, which can only be
used from PPC side. They have some advantages over hunk libs.
Or, you can have an ELF lib with an 68k-style jump table and call that
using proper stubs or macros. This can be used to create 68k-style devices,
although I probably would do it with ports..
--
// http://www.lut.fi/~tsuikki/ // 22-line signature compressed to 10%:
char*s,p[911],*t="=yE)8A*I3U-}{=@EHW8!=7@TT23sESTJHqO--OOki~zuik";main(){for(
memset(p,39,99);p[*t++]^=6;p[*t++]^=8)for(puts(s=p+25);*s|=s[1]&9;*++s&=*p);}
""msaupe@saale-net.de
>Therefore I don't see any legitimation to defend Phase5's Anti-Warpup
>actions (with all these flash ROM problems, which easily could have
>been solved by Phase5) just based on their image as 'saver of the
>AMIGA'.
I never said it was OK to have that anti-warpup code in the flash.
I just can't accept if Steffen writes ppc.library software wouldn't
be Amiga software.
>I also don't see any reason, why all this still continues.
I don't have problems with WarpUp becoming a defacto standard, but
someone else has a problem that there were people coding for PowerUp.
--
- Marco Saupe - Team *AMIGA* - http://www.saale-net.de/homepages/m.saupe -
------ msa...@saale-net.de --- Marco...@in.stud.tu-ilmenau.de -------
------ AMIGA 1200T PPC 603/200 -------- Amiga Zentrum Thueringen -------
C:\ Bad command or file name! Go stand in the corner.
""msaupe@saale-net.de
> ""msa...@saale-net.de wrote :
>And this allows them to destroy the market ?
I don't say it's ok that the anti Warp code is in it.
But I can't agree if you say P5 stuff isn't Amiga.
You can criticize ppc.lib, but it's unfair to slag P5 off
this way. PPC.lib may be badly coded (I can't judge here),
the API may be different, whatever... but if you reduce it
to the main fact then Warp and PPC.lib are both extensions to
the *Amiga*OS to make use of the PPC.
>""> But the PowerUp programs that exist won't stop working (we can talk
>""> about the Escena board when they arrive) and they are Amiga programs.
>They will on Escena Board. At least most of them (as only around 10% run on
>the ppc.lib Emulation)
I wouldn't have said anything if you had suggested that all new
PPC programs should be done for WarpUp as it will be the standard.
I wouldn't have said anything if you had suggested that the coders of
PowerUp supporting programs should port them to WarpUp.
But I can't accept if you say PowerUp is no Amiga.
AmigaBasic is from Microsoft, it doesn't run on newer machines, but
it's *Amiga*Basic. Viirii are a bad thing (TM) but the fact they
"run" on the Amiga makes them Amiga-Viirii.
And don't misunderstand those examples of why bad things are still
Amiga-stuff -> I don't want to say that ppc.lib is bad (I can't judge)
or even that it would be a Virus; I just want to make clear that the
fact it might be worse than WarpUp or that some things may be different
doesn't make it "not Amiga".
>Amiga-Developpers, it means at the same time "ppc.lib stuff" and "bad coded
>software", you have to get the rest of the sentence to find out which of
>them was meant this time...) (Well, seriously... for ppc.library you bought
>it...)
Fine then, so I'll run it on the ppc.library-OS as there's no raytracer
with comparable speed available for the Amiga...
- Marco Saupe - Team *AMIGA* - http://www.saale-net.de/homepages/m.saupe -
------ msa...@saale-net.de --- Marco...@in.stud.tu-ilmenau.de -------
------ AMIGA 1200T PPC 603/200 -------- Amiga Zentrum Thueringen -------
What happens to the hole when the cheese is gone?
Nathan Wain
On 6 Jan 1999 msa...@saale-net.de wrote:
> Because some people see that without them we wouldn't have PPC at all!
> And I highly doubt that another company would have done a PPC board;
> don't forget that without the P5 boards there wouldn't be WarpUp
> for the Escena board, yet.
Yeah, nice of them to do a PPC board. Really *really* shitty of them
to sabotage everyone's attempts to make any software for them.
> Even if WarpUp would be 1000 times better, you can't say ppc.library
> is no Amiga.
Wait for the Amiga's with no 68k processor... no powerup stuff
will work on that. And they were just plain lying about their
intentions to do a PPC-only system of their own.
> WarpUp will very probably be in OS 3.5 - Okay, good.
> Most new PPC programs will be for WarpUp - No problem
> But the PowerUp programs that exist won't stop working (we can
> talk about the Escena board when they arrive)
That's very nice and short-sighted of you... not unlike the
company you worship.
> and they are Amiga programs.
Not for long... which is the problem.
If you don't mind being conned out of money for some short-term
half-arsed solutions by the most microsoft-like company on the
Amiga side of the pond, fine. I prefer a little thought with
my rationale though thanks.
Nathan.
Steffen Haeuser
""msa...@saale-net.de wrote :
Hi!
""> I don't say it's ok that the anti Warp code is in it.
""> But I can't agree if you say P5 stuff isn't Amiga.
""> You can criticize ppc.lib, but it's unfair to slag P5 off
""> this way. PPC.lib may be badly coded (I can't judge here),
""> the API may be different, whatever... but if you reduce it
Yes, it is VERY different from all that was done on Amiga up to now.
""> I wouldn't have said anything if you had suggested that all new
""> PPC programs should be done for WarpUp as it will be the standard.
""> I wouldn't have said anything if you had suggested that the coders of
""> PowerUp supporting programs should port them to WarpUp.
""> But I can't accept if you say PowerUp is no Amiga.
I accept your point. But i still keep my opinion :)
""> Fine then, so I'll run it on the ppc.library-OS as there's no raytracer
""> with comparable speed available for the Amiga...
:)
Steffen
Steffen Haeuser
tsu...@feandil.lnet.lut.fi wrote :
Hi!
>>- I connect AmigaOS also very close with it's hunk format. It is a
>> resource-saving SMALL format, while ELF is big and bloated. Weren't we
ts> Duh. Hunk format is designed with 68k in mind, has known bugs and doesn't
ts> support the split relocation needed for PPC code.
It can be slightly expanded in a still-compatible way.
ts> ELF format is universal, quite clean format, and supports anything needed
ts> and much more. PPC programs don't run on 68k processor anyway, so why would
ts> it be necessary to use a 68k file format?
ELF is a monster !!!
ts> ELF files are slighly bigger than hunk files, but "big and bloated" is
ts> quite an over-estimation.. I compiled some test programs with scppc using
ts> hunk and ELF formats, and there's only a few kilobytes difference.
It is not. I also compiled some things, and there was sometimes factor 3 or 4
(EGCS for ELF).
>>- Look at the API. For WarpUP really only some "PPC" was appended to the names,
>> asides from that it looks 1:1 like the original API. Not true for ppc.lib
ts> This shows that you _never_ have even take a look at the ppc.library API..
ts> Here's a short list of ppc.library functions which only have "PPC"
ts> appended:
I know the API. Like my recent post on the ppc.lib Emulation proves. But for
example what ppc.lib calls a library and what AmigaOS calls a library is
different...
ts> I only included the names which were exact matches and which I was sure of,
ts> there are many "similar" functions like PPCGetMessage vs. GetMsg...
Exactly that what was i was pointing out. Why using a "new" API when there is
a existing one ? Why change the functionality of GetMsg ?
>>- No possibility to integrate it to 68k-OS by use of a 68k Emulator
ts> That's wrong. Both warpup and ppc.library implement 68k functions the same
ts> way; you call a kernel function which then runs the 68k code. On current
ts> systems this is done by the 68k processor.
I did not mean this. Try integrating a 68k Emulation in - current !!! -
ppc.library. You will have problems, as ppc.lib does in some aspects require a
68k Kernel "below". WarpUP does not require a 68k Kernel, it can run 100% PPC
Native.
Steffen Haeuser
""msaupe@saale-net.de
>> WarpUp will very probably be in OS 3.5 - Okay, good.
>> Most new PPC programs will be for WarpUp - No problem
>> But the PowerUp programs that exist won't stop working (we can
>> talk about the Escena board when they arrive)
>That's very nice and short-sighted of you... not unlike the
>company you worship.
There's nothing wrong with porting them to Warp to make the run on the
Escena boards (if they arrive). Yes, do it. The more PPC programs the
better. But that doesn't make the ppc.lib programs "not amiga"
I don't worship them. I also don't like the move with the anti-warp code.
But unlike others I compare their good and their bad moves and think
there are still more good ones.
>> and they are Amiga programs.
>Not for long... which is the problem.
Well, Carrier Command doesn't work on my 1200. Does that mean that
it isn't an Amiga program? I don't say PowerUp is better than Warp
(both work good from the users POV and can't say about the developers
POV) but it doesn't make it "not amiga".
That's the point no one here seems to understand.
--
Jyrki O Saarinen
> - I connect AmigaOS also very close with it's hunk format. It is a
> resource-saving SMALL format, while ELF is big and bloated. Weren't we
> always laughing about the PC-guys with their huge memory-needs and that ?
> With ppc.lib it is their turn to laugh back :)
Please, how the hell an executable format is able to bloat executables?
Perhaps somehow magically doubling ppc executable size by making
instruction word length 8 bytes instead of 4?
Jyrki O Saarinen
> Therefore I don't see any legitimation to defend Phase5's Anti-Warpup
> actions (with all these flash ROM problems, which easily could have
Why wouldn't they protect their product? After all, you created
a double standard which confuses potential ppc cards buyers.
(which is better or worse or anything doesn't affect this matter at all)
Jyrki O Saarinen
> ELF is a monster !!!
Why not say ELF is anti-christ too?
> It is not. I also compiled some things, and there was sometimes factor
> 3 or 4 (EGCS for ELF).
Just because you are too stupid to strip debug information out of the
executable, don't make false assumptations.
Sam Jordan
> > Therefore I don't see any legitimation to defend Phase5's Anti-Warpup
> > actions (with all these flash ROM problems, which easily could have
>
> Why wouldn't they protect their product? After all, you created
Protect by sabotage?
And this argument might be valid when WarpUp came out (because at
that time, the chance was there, that WarpUp actually would not
provide a sufficient quality), but today it's useless. WarpUp is
a valid alternative to PowerUp, therefore there is no need to
protect the P5 hardware product. Now, the problems with WarpUp
and P5 hardware actually decrease the worth of the P5 hardware
products, since the amount of reliable working PPC software is
smaller as when WarpUp wouldn't have any problems. The more
PPC software is available the more attractive the PPC boards
will become. Many people simply don't buy PPC accelerators because
of lack of available PPC software.
> a double standard which confuses potential ppc cards buyers.
The confusion could have been *MUCH* reduced, if they didn't do any
kind of sabotage. It would have been easily possible to develop together
easy installation methods to allow the user to choose between systems.
The integration of the PPC could have been much better, and the users
would be happy, there would be no termination, no reliability problems.
After all we were always for the freedom of choice, we never wanted
to force anyone to use our system (the fact, that the users actually
have to use WarpOS on newer PPC-only cards is that P5 simply doesn't
provide
a working software solution for this kind of architecture). I don't see
anything problematic in this philosophy. I just don't see why people
agree with P5's philosophy of attempting to force the user to use their
system, especially since the freedom of choice is perfectly available
for CSPPC users, so why not for the BPPC users?
Would you like to be forced to do something, when you could have the
free choice?
bye
--
Sam Jordan
Thomas Johnsrud
fl>>>> And get the PowerUP version that is much better and newer.
>>>We will talk again when WarpUP runs on the Escena-Board, while ppc.library
>>>version still on the PowerUP one :)
>>>And yes, i will adapt the latest SNES9x sources into it later... currently
>>>i am busy with a version of ADescent which supports BOTH WarpUP and Warp3D
>>>!!!
>>>BTW: The only case where ppc.library version of SNES is better is on AGA
>>>crap... on GFX Board Systems my version is faster...
TJ>> Bah! On my machine the WarpUp version, always crashes!
> On my machine too. Only H&P's demos work.
Jupp, I have heard several other pepole say it crashes to.
Teemu Suikki
>ts> hunk and ELF formats, and there's only a few kilobytes difference.
>It is not. I also compiled some things, and there was sometimes factor 3 or 4
>(EGCS for ELF).
You do know how to strip your binaries, don't you?
I can make my hunk binaries 3 or 4 times larger than ELF, if I compile with
DEBUG=FULL...
>ts> there are many "similar" functions like PPCGetMessage vs. GetMsg...
>Exactly that what was i was pointing out. Why using a "new" API when there is
>a existing one ? Why change the functionality of GetMsg ?
Because GetMsg doesn't work on multiprocessor system. Exec messaging uses
locking with Forbid and Permit, which doesn't prevent the PPC from entering
the same code at the same time... So, if you want a working PPCGetMsg, you
must run it in 68k code which is slow.
If you need to use standard 68k ports in powerup, you _can_ use normal exec
GetMsg... It's just that PPCGetMessage runs PPC native and GetMsg doesn't,
so for obvious reasons PPCGetMessage is preferred.
>I did not mean this. Try integrating a 68k Emulation in - current !!! -
>ppc.library. You will have problems, as ppc.lib does in some aspects require a
>68k Kernel "below". WarpUP does not require a 68k Kernel, it can run 100% PPC
Argh. I'm 100% sure that the _current_ warpup doesn't work in PPC-only
machine. It doesn't even have the emulation yet, how could it work?
Both systems can work with 68k emulation, if the necessary changes are
made. API already supports this in both systems.
Steffen Haeuser
flek...@online.no wrote :
Hi!
>> On my machine too. Only H&P's demos work.
fl> Jupp, I have heard several other pepole say it crashes to.
Well, this is - like i said in this thread already SEVERAL times - the problem
of the LoadSeg-Patch. A PPC Shared Library requires LoadSeg to be called, and
when termination is not correctly performed on A1200 PPC this causes a crash
due to the "Anti-WarpUP Code" (the patch assumes that ppc.lib can always be
opened, and does not check the return value of OpenLibrary before executing a
ppc.lib function).
This means of course, that WarpUP programs which do not call LoadSeg for
Libloading - like the WarpUP demos - don't have any problem. But programs
which DO use such libs (for example rtgmaster) *have* deep trouble. Well, with
the old version of the "Anti-WarpUP Code" it did not happen like this...
Well, and i certainly won't stop using PPC Shared Libs because of this. P5
should fix it by removing this damned patch IMHO, and best put the ppc.lib to
libs: instead of the ROM, like on the CyberStorm PPC...
Well, if you install Terminator correctly - and don't activate ANY ppc.lib
program before starting WarpUP - including CGX PPC and PPCInstall - then those
problems don't appear...
Steffen Haeuser
Steffen Haeuser
jxsa...@cs.Helsinki.FI wrote :
Hi!
jx> Steffen Haeuser <mag...@birdland.es.bawue.de> wrote:
>> ELF is a monster !!!
jx> Why not say ELF is anti-christ too?
As this is obvious i did not mention it :)
>> It is not. I also compiled some things, and there was sometimes factor
>> 3 or 4 (EGCS for ELF).
jx> Just because you are too stupid to strip debug information out of the
jx> executable, don't make false assumptations.
Don't get insulting... well, i always compared executable sizes. I did not
compare SAS/C ELF <-> SAS/C Hunk, but instead EGCS ELF <-> StormC Hunk. And
well, this is not only as to my own programs. Compare the
PPC-tools-and-packer-stuff which me and Andreas Kleinert ported... mine are
always definitely smaller executables... (unless you want to say that Andreas
Kleinert would also be "too stupid"... i think as to ppc.library he is a
proficient programmer...)
Steffen Haeuser
Steffen Haeuser
jxsa...@cs.Helsinki.FI wrote :
Hi!
>> - I connect AmigaOS also very close with it's hunk format. It is a
>> resource-saving SMALL format, while ELF is big and bloated. Weren't we
>> always laughing about the PC-guys with their huge memory-needs and that ?
>> With ppc.lib it is their turn to laugh back :)
jx> Please, how the hell an executable format is able to bloat executables?
jx> Perhaps somehow magically doubling ppc executable size by making
jx> instruction word length 8 bytes instead of 4?
I don't know HOW, but it happens... just compare some ppc.lib programs to some
WarpUP versions of them.
One more example: MAME is around 8 MB for ppc.library... my version for WarpUP
(it is not yet working, sadly, but all files are already compiling) is about
4.5 MB...
Steffen Haeuser
Jyrki O Saarinen
> provide a sufficient quality), but today it's useless. WarpUp is
> a valid alternative to PowerUp, therefore there is no need to
> protect the P5 hardware product. Now, the problems with WarpUp
Protect or not, I don't care but the fact is that you created
a double standard which confuses people and most importanly,
developers.
Why a double standard had to be created in the first place?
Sure, ppc.library wasn't very far in 1996 but which software
is perfect in it's early days? None.
> After all we were always for the freedom of choice, we never wanted
> to force anyone to use our system (the fact, that the users actually
> have to use WarpOS on newer PPC-only cards is that P5 simply doesn't
> provide
The possible future ppc cards (which I find quite unlikely, but nice
if they arrive) are irrelevant in this discussion.
Not forcing anyone to use - well how about the lack of free
production quality compiler? (not to mention ceratain Mr. H.
yelling 'use WarpOS or die')
If I would have planned a major ppc software development earlier,
the choices would have been vbcc (not production quality at least in
the WarpOS version) and StormC which maybe good these days, but
earlier ... well. For example, see the benchmarks compiled and
ran by Carsten S. (I don't remember his last name, but he is the guy
who made the custom 68060.library), many of the _small_ benchmarks
compiled on StormC 2.x either crash or give wrong results. This
doesn't give a very reliable image of a product.
This was in 1997. Now even if 3.x would be perfect, this kind of
history can't not be affecting the decision as you surely can understand.
> anything problematic in this philosophy. I just don't see why people
> agree with P5's philosophy of attempting to force the user to use their
> system, especially since the freedom of choice is perfectly available
> for CSPPC users, so why not for the BPPC users?
I agree.
Jyrki O Saarinen
> Wait for the Amiga's with no 68k processor... no powerup stuff
> will work on that. And they were just plain lying about their
> intentions to do a PPC-only system of their own.
Escena boards will need a 68k in the CPU-slot for deliviring
interrupts from the Amiga motherboards to ppc in the Z3 slot.
Sam Jordan
> > provide a sufficient quality), but today it's useless. WarpUp is
> > a valid alternative to PowerUp, therefore there is no need to
> > protect the P5 hardware product. Now, the problems with WarpUp
>
> Protect or not, I don't care but the fact is that you created
> a double standard which confuses people and most importanly,
> developers.
I'm repeating again, that the confusion could have been *MUCH*
reduced, if P5 didn't fight WarpUp. We didn't want to confuse
anybody, we just wanted to provide another product of the same
category, like it happens often in sw industry.
> > After all we were always for the freedom of choice, we never wanted
> > to force anyone to use our system (the fact, that the users actually
> > have to use WarpOS on newer PPC-only cards is that P5 simply doesn't
> > provide
>
> The possible future ppc cards (which I find quite unlikely, but nice
Two projects are in progress, I can't see why it should be unlikely,
that at least one of these projects will appear.
> Not forcing anyone to use - well how about the lack of free
> production quality compiler? (not to mention ceratain Mr. H.
Everyone can write a compiler for WarpUp, all informations are
available. It's not H&P's problem, when there are none. WarpUp
came out more than a year ago, time enough to adapt the compilers.
Btw, the egcs adaption will probably cause major problems with
WarpUp, since it seems, that it still works with V4-Stackframes.
Most library functions store r2 to 20(r1), where application
data might be located in V4-Applications. More problems are
to be expected for functions with many parameters. Therefore
the egcs adaption will probably only be usable if someone adds
PowerOpen support to it (or some kind of stackframe conversion
for library calls).
> If I would have planned a major ppc software development earlier,
> the choices would have been vbcc (not production quality at least in
> the WarpOS version) and StormC which maybe good these days, but
> earlier ... well. For example, see the benchmarks compiled and
That's the same with all software. You very often bring the
example of the ppc.library and ask us why we didn't want to
use this early ppc.library but decided to use something else.
Now I can also ask you, why you didn't want to work with early
StormC? You usually bring the argument concerning ppc.library,
that H&P should have waited until ppc.library would be
improved in quality. Finally H&P had to wait for almost a year,
until the ppc.library reached a usable state. Now we can turn
the focus again and I can ask you, why you didn't just wait
until StormC gets better? Of course, you didn't want to wait,
but choosed another option, that's reasonable.
Basically I suppose, that we both had similar reasons:
the quality was unsufficient at these days. I hope that this
helps to understand why we decided to go for our own PPC
solution, just as you decided to use a compiler, which provided
the quality you needed.
bye
--
Sam Jordan
Nathan Wain
On 7 Jan 1999 msa...@saale-net.de wrote:
> Nathan rote:
>
> >> WarpUp will very probably be in OS 3.5 - Okay, good.
> >> Most new PPC programs will be for WarpUp - No problem
> >> But the PowerUp programs that exist won't stop working (we can
> >> talk about the Escena board when they arrive)
>
> >That's very nice and short-sighted of you... not unlike the
> >company you worship.
>
> There's nothing wrong with porting them to Warp to make the run on the
> Escena boards (if they arrive). Yes, do it. The more PPC programs the
> better. But that doesn't make the ppc.lib programs "not amiga"
We are rapidly losing programmers for this (currently) dying platform.
To say that it's okay to program for PowerUp now, because it can be
ported to WarpUp later, assumes that the programmer will still be
around later (and can be bothered porting it), or that the source-code
is available (and someone else can be bothered porting it.)
Whether PowerUp should be considered "not Amiga" or not is very
subjective... But since it's buggy, a kludge, and not future-proofed,
I think that's enough reason to stop developing for it *now*. (IMHO)
> I don't worship them. I also don't like the move with the anti-warp code.
> But unlike others I compare their good and their bad moves and think
> there are still more good ones.
Really? Hmmmm...
> >> and they are Amiga programs.
>
> >Not for long... which is the problem.
>
> Well, Carrier Command doesn't work on my 1200. Does that mean that
> it isn't an Amiga program?
Entirely different situation back then...
There was no alternative to a 7.5MHz A500/1000 back then, (and no
non-68000 Amiga soon to appear) and the early programmers were still
in the 8bit hit-the-hardware mode of thinking.
Also, use of Carrier Command didn't stop half the community using
"Warp Command", and entirely better game, which would work on the new
A1200's the community knew was going to appear soon.
(BTW, Carrier command does work nicely on my A1200.) :-)
> I don't say PowerUp is better than Warp
> (both work good from the users POV and can't say about the developers
Try a few heavy-processing tricks to see the advantage of dynamic
task scheduling, you'll soon see which works better.
> POV) but it doesn't make it "not amiga".
> That's the point no one here seems to understand.
Oh, I can understand it. Just don't expect me to agree with it.
Nathan.
Nathan Wain
On 7 Jan 1999, Jyrki O Saarinen wrote:
> Sam Jordan <s...@spin.ch> wrote:
> > Therefore I don't see any legitimation to defend Phase5's Anti-Warpup
> > actions (with all these flash ROM problems, which easily could have
>
> Why wouldn't they protect their product?
Hey, they *already* had the hardware market sealed up... And that's
where they were making all the money. If they had had any intelligence
they would've seen that the Amiga community was too small to split up
about the underlying software.
> After all, you created
> a double standard which confuses potential ppc cards buyers.
Yeah, but it wasn't Haage and Partner that made using WarpUp an all
or nothing affair. But anyway, we already know how much you hate
them, so you don't have to show some twisted logic to justify that
opinion.
And now Phase 5 don't have any decent programmers left (not that
they had many before, mind) we would've been left with a dead standard
anyway.
> (which is better or worse or anything doesn't affect this matter at all)
Oh, of course it doesn't. (sarcasm)
I would've just loved to be stuck with a s**t standard. To *really*
believe they could create a PPC only system (the lying bastards) and
that they had the software-engineering skills to make PowerUp work on
it. No Phase 5 never screwed us over at all.
Nathan.
Nathan Wain
On 7 Jan 1999, Jyrki O Saarinen wrote:
No, you're mistaken... That's Phase 5. (A sub-cult of M.Soft BTW.) :-)
Michael Merkel
about <Re: Snes9x (PowerUP) kicks Warpsnes ass!>:
> WarpSNES is also based on 1.10. At least the latest version.
and where can we find the latest version, please?
byebye...
--
Michael Merkel (MiMe@IRC)
Michael...@gmx.net
http://home.t-online.de/home/Merkel.Michael/
Jyrki O Saarinen
>> Protect or not, I don't care but the fact is that you created
>> a double standard which confuses people and most importanly,
>> developers.
> I'm repeating again, that the confusion could have been *MUCH*
> reduced, if P5 didn't fight WarpUp. We didn't want to confuse
Sure.
Now, if you were p5 wouldn't you fight because kernel number two makes
selling your ppc cards harder. Can you see the point?
> anybody, we just wanted to provide another product of the same
> category, like it happens often in sw industry.
Good idea but in real world this turned out to be not-so-good in
general.
Maybe things would have been otherwise if the product you offered
would have been 'a normal application'.
>> The possible future ppc cards (which I find quite unlikely, but nice
>
> Two projects are in progress, I can't see why it should be unlikely,
> that at least one of these projects will appear.
What is the 2nd one besides Escena?
> Everyone can write a compiler for WarpUp, all informations are
> available. It's not H&P's problem, when there are none. WarpUp
> came out more than a year ago, time enough to adapt the compilers.
A funny attitude IMHO.
You create a new executable format while there would have been formats
with prooven support ready and available, and then say it is not H&P's
problem if there are no dev tools?
I can understand creating own object code format boost StormC sales a
bit but that can't be that much these days.
Enough time to adapt dev tools, true. But why people would have done
it since there was already elf tools available, ready and proven to
work correctly?
> Btw, the egcs adaption will probably cause major problems with
> WarpUp, since it seems, that it still works with V4-Stackframes.
Yes it is nothing more than a new startup code + elf loader if I
understod correctly.
>> If I would have planned a major ppc software development earlier,
>> the choices would have been vbcc (not production quality at least in
>> the WarpOS version) and StormC which maybe good these days, but
>> earlier ... well. For example, see the benchmarks compiled and
> That's the same with all software. You very often bring the
> example of the ppc.library and ask us why we didn't want to
> use this early ppc.library but decided to use something else.
In 1996 ppc.library wasn't as stable as it is today, so you
created a custom system instead of finding bugs and reporting them?
> Now I can also ask you, why you didn't want to work with early
> StormC? You usually bring the argument concerning ppc.library,
> that H&P should have waited until ppc.library would be
> improved in quality. Finally H&P had to wait for almost a year,
> until the ppc.library reached a usable state. Now we can turn
> the focus again and I can ask you, why you didn't just wait
> until StormC gets better? Of course, you didn't want to wait,
> but choosed another option, that's reasonable.
Now, compaing a kernel and a compiler is a bit different;
you rarely need to use ppc kernel functions in an application.
But you need a compiler producing correct code to make
your application. There is difference.
> the quality was unsufficient at these days. I hope that this
> helps to understand why we decided to go for our own PPC
> solution, just as you decided to use a compiler, which provided
Partly yes. Not completely because you can't compare a compiler
producing broken code and a inmature kernel.
PS. It's good to discuss this with someone who doesn't spread
disinformation and half-thruts. You are able to discuss this
at correct level while a certain german can't.
Jyrki O Saarinen
> where they were making all the money. If they had had any intelligence
> they would've seen that the Amiga community was too small to split up
> about the underlying software.
Now, think again which product splitted the ppc software field.
It surely wasn't ppc.library because it came first.
This is the point I'm getting at.
> Yeah, but it wasn't Haage and Partner that made using WarpUp an all
> or nothing affair. But anyway, we already know how much you hate
> them, so you don't have to show some twisted logic to justify that
> opinion.
Facts and relevant points instead of personal shit please.
But that's what most people do when they ran out of arguments, get personal.
> And now Phase 5 don't have any decent programmers left (not that
> they had many before, mind) we would've been left with a dead standard
> anyway.
I have the feeling p5 never had any programmers on their
payroll, all work done 'out house'.
H&P has what two people (Sam & Michael) updating a C/C++ compiler,
ppc kernel, OpenGL implementation ...? Mr Jordan, correct here please.
About your last chapter; do you happen to believe in worldwide
conspiracy in UFOs and Area51 too?
Jyrki O Saarinen
> Whether PowerUp should be considered "not Amiga" or not is very
> subjective... But since it's buggy, a kludge, and not future-proofed,
> I think that's enough reason to stop developing for it *now*. (IMHO)
Buggy? What function is buggy and under what conditions this bug occurs?
Kludge? The whole PPC in an Amiga is a huge kludge.
> Try a few heavy-processing tricks to see the advantage of dynamic
> task scheduling, you'll soon see which works better.
You have constant X amount of cpu cycles in a second. Having a
scheduler which has dynamic priorities doesn't increase the
amount of those cpu cycles, just how they are divided
between different processes (tasks, threads).
Nicholai Benalal
SJ> example of the ppc.library and ask us why we didn't want to use
SJ> this early ppc.library but decided to use something else. Now I
SJ> can also ask you, why you didn't want to work with early StormC?
SJ> You usually bring the argument concerning ppc.library, that H&P
SJ> should have waited until ppc.library would be improved in
SJ> quality. Finally H&P had to wait for almost a year, until the
SJ> ppc.library reached a usable state. Now we can turn the focus
SJ> again and I can ask you, why you didn't just wait until StormC
SJ> gets better? Of course, you didn't want to wait, but choosed
SJ> another option, that's reasonable.
SJ>
SJ> Basically I suppose, that we both had similar reasons: the
SJ> quality was unsufficient at these days. I hope that this helps to
SJ> understand why we decided to go for our own PPC solution, just as
SJ> you decided to use a compiler, which provided the quality you
SJ> needed.
SJ>
That comparison is very weak imho. You can run two programs compiled
with different compilers at the same time if the two compilers are
targeted at the same kernel. If the executable files are for
different kernels, it's very hard to make them work in parallell.
For the enduser, it's of no interest what compiler was used...but
it's very important what kernel he/she is supposed to use to run the
application. It's a much bigger step to use/develop a different
kernel than to choose another compiler.
/Nicholai
Kresimir Farkas
>CSPPC users don't have problems with WarpUp, except if they
>did something wrong and didn't want to mail me to ask for
>solutions (I'm usually responding immediately after mails
>come in).
I have. WarpUp crashes 9 out of 10 times when terminating ppc.library.
Even after cold reboot, and nothing executed on the PPC.
Faki
----
Kresimir Farkas
>Still it has not much to do anymore with the Amiga :)
>(And crashes for many people's systems...)
On my Amiga only WarpUp crashes.
°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°
° Faki - A4000/040-25+604e-180/66 MB/6 Gb/8x CD/PIV/MFC III °
°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°
Rocou
SH> flek...@online.no wrote :
SH> Hi!
>>> On my machine too. Only H&P's demos work.
fl>> Jupp, I have heard several other pepole say it crashes to.
SH> Well, this is - like i said in this thread already SEVERAL times - the
SH> problem of the LoadSeg-Patch.
[...]
SH> Well, if you install Terminator correctly - and don't activate ANY
SH> ppc.lib program before starting WarpUP - including CGX PPC and
SH> PPCInstall - then those problems don't appear...
OK but why H&P's demos work fine?
Steffen Haeuser
Michael...@gmx.net wrote :
Hi!
><mag...@birdland.es.bawue.de> wrote at 05 Jan 1999 06:54:59
Mi> about <Re: Snes9x (PowerUP) kicks Warpsnes ass!>:
>> WarpSNES is also based on 1.10. At least the latest version.
Mi> and where can we find the latest version, please?
Should be on Aminet. If not, try come.to/WarpSNES or www.classicgaming.com/aec
Steffen Haeuser
Steffen Haeuser
nat...@caverock.net.nz wrote :
Hi!
na> On 7 Jan 1999, Jyrki O Saarinen wrote:
>> Steffen Haeuser <mag...@birdland.es.bawue.de> wrote:
>>
>> > ELF is a monster !!!
>>
>> Why not say ELF is anti-christ too?
na> No, you're mistaken... That's Phase 5. (A sub-cult of M.Soft BTW.) :-)
Well, i would not put it that way... well, if only they would have given
Software-Developpement completely to H&P at the start... then the current MESS
on the market would not have happened... and their software-developper(s)
could have concentrated on other things, also...
Steffen Haeuser
Steffen Haeuser
jxsa...@cs.Helsinki.FI wrote :
Hi!
jx> Now, if you were p5 wouldn't you fight because kernel number two makes
jx> selling your ppc cards harder. Can you see the point?
Well, it was in talk at the start, that they gave software developpement
completely to H&P, dropping ppc.library... this would have been good, if they
would have done this at the start ... then the current problems would never
have appears. And the big problem of P5 *always* was the software. If they
could let their software developpers work on other things, maybe then we even
would have 3D Drivers for Permedia 2 already ?
jx> Maybe things would have been otherwise if the product you offered
jx> would have been 'a normal application'.
But this would not solve the problems of the lib...
>>> The possible future ppc cards (which I find quite unlikely, but nice
>>
>> Two projects are in progress, I can't see why it should be unlikely,
>> that at least one of these projects will appear.
jx> What is the 2nd one besides Escena?
I think if it could be already named, Sam would surely have named it ... This
is the same discussion like about Emulator-Speed in the other newsgroup. When
there is something which can be released, it WILL be released surely (and no,
i don't know what names Sam has in mind...).
jx> A funny attitude IMHO.
jx> You create a new executable format while there would have been formats
Amiga-Hunk Format is no new executable format (as you can read in the Docs,
there is NO difference in the format of a 68k Amiga Executable and a WarpUP
one... there are slight differences - a few new Hunk-Types - in the OBJECT
format, but NOT in the executable format...).
jx> I can understand creating own object code format boost StormC sales a
jx> bit but that can't be that much these days.
This was definitely not the reason. In fact if they would have wanted, they
could have done StormC for ppc.lib... but out of other reasons that would not
have been good.
jx> Enough time to adapt dev tools, true. But why people would have done
jx> it since there was already elf tools available, ready and proven to
jx> work correctly?
I remind you: StormC PPC was available much before SAS/C PPC and EGCS PPC !!!
>> Btw, the egcs adaption will probably cause major problems with
>> WarpUp, since it seems, that it still works with V4-Stackframes.
jx> Yes it is nothing more than a new startup code + elf loader if I
jx> understod correctly.
Yes, sadly. But at least it's executables can run parallel with "normal"
WarpUP ones...
jx> PS. It's good to discuss this with someone who doesn't spread
jx> disinformation and half-thruts. You are able to discuss this
jx> at correct level while a certain german can't.
Well, i only react on things other people say. :)
Steffen Haeuser
Steffen Haeuser
jxsa...@cs.Helsinki.FI wrote :
Hi!
jx> production quality compiler? (not to mention ceratain Mr. H.
jx> yelling 'use WarpOS or die')
Well, i did not say this. Also: You always had the attitude 'don't use StormC
or die'. Who is the fanatic now ?
Steffen Haeuser
Steffen Haeuser
jxsa...@cs.Helsinki.FI wrote :
Hi!
>> Wait for the Amiga's with no 68k processor... no powerup stuff
>> will work on that. And they were just plain lying about their
>> intentions to do a PPC-only system of their own.
jx> Escena boards will need a 68k in the CPU-slot for deliviring
jx> interrupts from the Amiga motherboards to ppc in the Z3 slot.
True. But this is no time-intensive stuff. The point is that OS-calls will
speedup MUCH. The OS will - with exception of interrupts - run PPC Native.
Steffen Haeuser
Steffen Haeuser
tsu...@feandil.lnet.lut.fi wrote :
Hi!
ts> You do know how to strip your binaries, don't you?
ts> I can make my hunk binaries 3 or 4 times larger than ELF, if I compile with
ts> DEBUG=FULL...
Hmmm, this i did not manage. Even if my StormC compiled programs were with
Debug-info, ELF was still bigger...
>>a existing one ? Why change the functionality of GetMsg ?
ts> Because GetMsg doesn't work on multiprocessor system. Exec messaging uses
But then neither does ppc.lib...
ts> If you need to use standard 68k ports in powerup, you _can_ use normal exec
ts> GetMsg... It's just that PPCGetMessage runs PPC native and GetMsg doesn't,
ts> so for obvious reasons PPCGetMessage is preferred.
>>I did not mean this. Try integrating a 68k Emulation in - current !!! -
>>ppc.library. You will have problems, as ppc.lib does in some aspects require a
>>68k Kernel "below". WarpUP does not require a 68k Kernel, it can run 100% PPC
ts> Argh. I'm 100% sure that the _current_ warpup doesn't work in PPC-only
ts> machine. It doesn't even have the emulation yet, how could it work?
It has the Emulation. At least on my machine (i am betatester).
ts> Both systems can work with 68k emulation, if the necessary changes are
ts> made. API already supports this in both systems.
Argh, too.
"If the necessary changes are done"... Who would create a 68k Emu for ppc.lib
? You also would need access to the ppc.lib sources, BTW... as internally
there are a lot of dependencies to the 68k Kernel "Exec" (which was not the
case for WarpUP...)
Steffen Haeuser
Steffen Haeuser
s...@spin.ch wrote :
Hi!
sj> Btw, the egcs adaption will probably cause major problems with
sj> WarpUp, since it seems, that it still works with V4-Stackframes.
sj> Most library functions store r2 to 20(r1), where application
sj> data might be located in V4-Applications. More problems are
sj> to be expected for functions with many parameters. Therefore
sj> the egcs adaption will probably only be usable if someone adds
sj> PowerOpen support to it (or some kind of stackframe conversion
sj> for library calls).
I think the conversion is already done (at least for PPC Shared Libs), using
Macros. powerpc.library functions run fine with it, BTW. As to further
developpement: I discussed with some people about it, that - if the compiler
would be changed to support real PowerOpen - vlink could be adapted to link
the object files. As i understood it before the main reason for not using
PowerOpen was that the linker had to be rewritten completely. Well, we will
see...
Steffen Haeuser
Nathan Wain
On 8 Jan 1999, Jyrki O Saarinen wrote:
> Sam Jordan <s...@spin.ch> wrote:
> > I'm repeating again, that the confusion could have been *MUCH*
> > reduced, if P5 didn't fight WarpUp. We didn't want to confuse
>
> Sure.
> Now, if you were p5 wouldn't you fight because kernel number two makes
> selling your ppc cards harder. Can you see the point?
What?!? Was there an ounce of rational thought involved in the
construction of that sentence!??!
They both require PPC hardware to run on. And, ohmygosh!, there
just happens to be only one company that sell it: Phase 5!
If someone wants to have a PPC, they buy it from Phase 5, it has
PowerUp with it... they're happy.
If someone wants WarpUp, they buy it from Haage and Parntner,
*and* buy a PPC board from Phase 5... they're happy.
Either way, Phase 5 wins.
Then there's scenario number three:
Phase 5 puts malicious code in ROMs of hardware. People have
PPC software that no longer works. People and companies see
there are problems. Companies refrain from releasing software.
People refrain from buying cards - and if you want an example
of that, I'm one. No-ones happy. And phase 5 sell less hardware.
WarpUp didn't threaten sales of the hardware... WarpUp *needed*
the hardware.
Nathan.
Nathan Wain
On 8 Jan 1999, Jyrki O Saarinen wrote:
> > where they were making all the money. If they had had any intelligence
> > they would've seen that the Amiga community was too small to split up
> > about the underlying software.
>
> Now, think again which product splitted the ppc software field.
> It surely wasn't ppc.library because it came first.
> This is the point I'm getting at.
How did Haage and Partner's software split it?
Both companies software was obtainable (for the users) for no
additional financial outlay. Phase 5 still gets all the money,
because they have the only hardware. Developers get to *choose*
which system to develop under... The user gets to use both.
The ideal:
We all could've been one happy community, using the one set of
hardware... Software being produced for PUp and WUp, the users
not caring which is which because they've installed both (or,
more likely, WUp was installed as part of the installation of
their first WUp product, and PUp was installed with the card)
and half the time the users don't even know which type they're
running.
Phase 5 still raking in all the money for their hardware...
So why did they have to put in that bloody code to f**k up the
users?
And, if they're such a nice company... Why was the bloody code
*still* present after they've even agreed with Amiga Tech. to
remove it.
...arrogance abounds with Phase 5... they seem to have had
problems with co-operating with (off the top of my head), Haage
and Partner, ProDad, VillageTronic, and finally our beloved
Amiga Technologies. Of course this is never Phase 5's fault...
> > Yeah, but it wasn't Haage and Partner that made using WarpUp an all
> > or nothing affair. But anyway, we already know how much you hate
> > them, so you don't have to show some twisted logic to justify that
> > opinion.
>
> Facts and relevant points instead of personal shit please.
> But that's what most people do when they ran out of arguments, get personal.
Wasn't the top bit a relevant fact? They never did threaten to
code WarpUp in a way that would force you to remove PowerUp. It
was to be an alternative, which would co-exist on any PowerUp system.
Your statements all imply a hatred for Haage and Partner or their
actions. And much of the Logic you present seems to be, well,
illogical - hence the "twisted" comment.
And was the rest personal? I apologise if it was interpreted as
being so. None of it was a slight on you. Just the arguments you
present.
After the same arguments for the 50th time, subtlety and diplomacy
tends to suffer.
> > And now Phase 5 don't have any decent programmers left (not that
> > they had many before, mind) we would've been left with a dead standard
> > anyway.
>
> I have the feeling p5 never had any programmers on their
> payroll, all work done 'out house'.
Well, whether in house or not, they basically have one software
engineer left. (And I use the term software engineer in it's
loosest possible sense to describe the kludges he does.)
> H&P has what two people (Sam & Michael) updating a C/C++ compiler,
> ppc kernel, OpenGL implementation ...? Mr Jordan, correct here please.
Well, so long as they continue to appear to have half a clue, and
they don't just up and leave online discussions because others dare
to have opinions that differ... (spot the phase5 employee comparison.)
> About your last chapter; do you happen to believe in worldwide
> conspiracy in UFOs and Area51 too?
Oooh definitely. There are also some strange subliminal messages
being passed to our youth through the Pringles potato-chip adverts
on TV... I'm sure the destructive effects on our society from this
will be felt soon.
Nathan.
(Just because you're not paraniod, doesn't mean they're *not* after you.)
Michael Rock
<snip>
> > And now Phase 5 don't have any decent programmers left (not that
> > they had many before, mind) we would've been left with a dead standard
> > anyway.
>
> I have the feeling p5 never had any programmers on their
> payroll, all work done 'out house'.
Yes, and that seems to become a real problem now.
They do not have access to the sources, f.e.
Imagine what would happen if Laire was to be run over by a bus ?
> H&P has what two people (Sam & Michael) updating a C/C++ compiler,
> ppc kernel, OpenGL implementation ...? Mr Jordan, correct here please.
We are more than that. Fact is, only Sam and me are reading news and
poste here.
> About your last chapter; do you happen to believe in worldwide
> conspiracy in UFOs and Area51 too?
Having read the last chapter of the posting you refer to, i can only
agree with
Nathan there. But what does that have to do with UFOs and Area51 ?
--
Michael Rock, student of computer science, TU Braunschweig.
Also at
Haage & Partner PPC-RISC Compiler department.
contact me at
m.r...@haage-partner.com
Speed is our business, and business runs goood.
A3oooT 060@50 + 604e@200
Nathan Wain
> > Whether PowerUp should be considered "not Amiga" or not is very
> > subjective... But since it's buggy, a kludge, and not future-proofed,
> > I think that's enough reason to stop developing for it *now*. (IMHO)
>
> Buggy? What function is buggy and under what conditions this bug occurs?
Well, for PUp case mentioned in this group where the machine will
consistantly crash after an indeterminate amount of time, but runs
just fine with WUp... (ah, the most fun kind of bug to track...) :-)
The WUp complaints I continually hear here are "The Demoes work,
but nothing else does"... which is Phase5's termination code anyway.
Buggy was too specific a description to describe all this... but
none of the above problem is the WUp software...
> Kludge? The whole PPC in an Amiga is a huge kludge.
Well, I consider one kludge more-elegant than the other.
When one requires a load-seg-patch, while the other doesn't
because it still uses the standard Amiga hunk format; when one
requires the presence of the 68k processor, and lots of superfluous
communication between it and the PPC, while the other has a HAL, and
is to have 68k emulation; and when one has static task scheduling
while the other has a dynamic scheduler...
That's enough reason for me to consider one kludge better implemented.
> > Try a few heavy-processing tricks to see the advantage of dynamic
> > task scheduling, you'll soon see which works better.
>
> You have constant X amount of cpu cycles in a second. Having a
> scheduler which has dynamic priorities doesn't increase the
> amount of those cpu cycles, just how they are divided
> between different processes (tasks, threads).
Yes, the CPU cycles per second is constant... most sentient
creatures can grasp this fact. But being able to re-allocate
them to tasks that the user is currently interacting with,
focus on the active window etc., adds hugely to the reponsiveness
of the system to the user. (Often with little effect on other
tasks, even if time-consuming and processor intensive.)
Playing around with, say, "Executive", on the old 50MHz 030 here
gave a huge boost in the productivity of the system. And is a
good indicator of what a Unix-like task-scheduler can do.
No matter how fast a system is, a PNG encoder (for example) that
gives itself a high task priority, and monopolises the CPU, will
make the system unusable when running in a standard Amiga fixed-
task-priority environment.
Nathan.
Nathan Wain
On Sat, 9 Jan 1999, Michael Rock wrote:
> Imagine what would happen if Laire was to be run over by a bus ?
Now there's a thought!....
How 'bout if we all imagine that really really *hard* for a few
minutes. The combined focus of will might influence the universe
in rewarding ways. :-)
Heheheheheeee...
Nathan. :-)
PS:
> > About your last chapter; do you happen to believe in worldwide
> > conspiracy in UFOs and Area51 too?
>
> Having read the last chapter of the posting you refer to, i can only
> agree with Nathan there. But what does that have to do with UFOs and
> Area51 ?
If they _were_ related, it would bring so much closure to the great
questions of this world, don't you think? :-)
Steffen Haeuser
nat...@caverock.net.nz wrote :
Hi!
na> After the same arguments for the 50th time, subtlety and diplomacy
na> tends to suffer.
Yeah, i also have that problem :)
Steffen Haeuser
Steffen Haeuser
k...@easynet.fr wrote :
Hi!
ka> SH> Hi!
>>>> On my machine too. Only H&P's demos work.
>fl>> Jupp, I have heard several other pepole say it crashes to.
ka> SH> Well, this is - like i said in this thread already SEVERAL times - the
ka> SH> problem of the LoadSeg-Patch.
ka> [...]
ka> SH> Well, if you install Terminator correctly - and don't activate ANY
ka> SH> ppc.lib program before starting WarpUP - including CGX PPC and
ka> SH> PPCInstall - then those problems don't appear...
ka> OK but why H&P's demos work fine?
They don't call LoadSeg inside PPC-Code ... only this causes problems. Inside
PPC-Code ppc.library cannot be successfully opened, and as the LoadSeg-Patch
of Ralp Schmidt calls ppc.library functions WITHOUT checking if the library
COULD be opened, this causes the known problems.
But a Shared Library on AmigaOS - anyways if 68k or PPC - *NEEDS* LoadSeg to
be working, instead of crashing...
The terminator just "fixes" LoadSeg again. But LoadSeg of course only can be
patched if it is not currently active (if a Seg is currently loaded, it cannot
be patched). And if such programs like mentioned above are loaded, then it
cannot be patched, and termination fails.
Okay, i guess now enough of it.
Steffen Haeuser
Steffen Haeuser
jxsa...@cs.Helsinki.FI wrote :
Hi!
jx> I have the feeling p5 never had any programmers on their
jx> payroll, all work done 'out house'.
jx> H&P has what two people (Sam & Michael) updating a C/C++ compiler,
jx> ppc kernel, OpenGL implementation ...? Mr Jordan, correct here please.
They have Alex, Jens, Jochen, Michael and Sam. And also some "outside"
developpers (for example Warp3D was mainly done by Hans-Joerg and Thomas
Frieden, at least the 68k Part)
Steffen Haeuser
Steffen Haeuser
fa...@usa.net wrote :
Hi!
>fl>> I run the ppc.library all the time, there is no problem.
>>Still it has not much to do anymore with the Amiga :)
>>(And crashes for many people's systems...)
fa> On my Amiga only WarpUp crashes.
On mine only ppc.library.
As on yours it crashes during termination it is definitely the cause of the
"Anti-WarpUP-Code". The PPC-Part of WarpUP is not even active yet during
termination.
Steffen Haeuser
Teemu Suikki
>Well, it was in talk at the start, that they gave software developpement
>completely to H&P, dropping ppc.library... this would have been good, if they
Why would they have done that?
Let's call to microsoft.. "Hi, your Windows 98 sucks. I made an OS myself,
will you stop selling your software now so I can take over the market"
>jx> it since there was already elf tools available, ready and proven to
>I remind you: StormC PPC was available much before SAS/C PPC and EGCS PPC !!!
And GCC was available years before that. GCC was available way before
Phase5 even started designing the board.
--
// http://www.lut.fi/~tsuikki/ // 22-line signature compressed to 10%:
char*s,p[911],*t="=yE)8A*I3U-}{=@EHW8!=7@TT23sESTJHqO--OOki~zuik";main(){for(
memset(p,39,99);p[*t++]^=6;p[*t++]^=8)for(puts(s=p+25);*s|=s[1]&9;*++s&=*p);}
Teemu Suikki
>ts> Because GetMsg doesn't work on multiprocessor system. Exec messaging uses
>But then neither does ppc.lib...
Argh!! I meant PPC + 68060. And I did explain that in my earlier post, but
you again managed to ignore that. A PPC version of GetMsg would not work on
68k/ppc system, that was the point.
WarpUP doesn't work on multi-PPC system either, by the way. It's possible
that such version appears later, but currently it doesn't work.
>ts> Argh. I'm 100% sure that the _current_ warpup doesn't work in PPC-only
>ts> machine. It doesn't even have the emulation yet, how could it work?
>It has the Emulation. At least on my machine (i am betatester).
And you have PPC-only machine too? And, _current_ warpup doesn't have the
emulation.
>ts> Both systems can work with 68k emulation, if the necessary changes are
>ts> made. API already supports this in both systems.
>"If the necessary changes are done"... Who would create a 68k Emu for ppc.lib
What has this to do with anything? You claimed that ppc.library "can't work
with 68k emulation".. Warpup can't work with emulation either if no-one
writes one for it. :)
>? You also would need access to the ppc.lib sources, BTW... as internally
>there are a lot of dependencies to the 68k Kernel "Exec" (which was not the
>case for WarpUP...)
Once again you are mixing things. I'm not writing any future ppc.library so
I don't need the sources either. Any possible ppc.library updates are
coming from phase5, and I'm quite sure that they have sources and can
remove the exec dependencies when needed.
It will be a long time before any PPC-only machines become available, if
ever.. Escena board require 68k too, remember. Your point is quite
irrelevant..
Steffen Haeuser
m.r...@haage-partner.com wrote :
Hi!
>> About your last chapter; do you happen to believe in worldwide
>> conspiracy in UFOs and Area51 too?
m.> Having read the last chapter of the posting you refer to, i can only
m.> agree with
m.> Nathan there. But what does that have to do with UFOs and Area51 ?
He probably wanted to indicate that all people who don't agree with him have
to be mad :) Fanatic...
Steffen Haeuser
Steffen Haeuser
tsu...@feandil.lnet.lut.fi wrote :
Hi!
ts> Argh!! I meant PPC + 68060. And I did explain that in my earlier post, but
ts> you again managed to ignore that. A PPC version of GetMsg would not work on
ts> 68k/ppc system, that was the point.
But not mine.
ts> WarpUP doesn't work on multi-PPC system either, by the way. It's possible
ts> that such version appears later, but currently it doesn't work.
I never claimed this.
>>ts> Argh. I'm 100% sure that the _current_ warpup doesn't work in PPC-only
>>ts> machine. It doesn't even have the emulation yet, how could it work?
>>It has the Emulation. At least on my machine (i am betatester).
ts> And you have PPC-only machine too? And, _current_ warpup doesn't have the
ts> emulation.
Well, if i run the Emulation the 68k is only used for Interrupt-Handling...
>>ts> Both systems can work with 68k emulation, if the necessary changes are
>>ts> made. API already supports this in both systems.
>>"If the necessary changes are done"... Who would create a 68k Emu for ppc.lib
ts> What has this to do with anything? You claimed that ppc.library "can't work
ts> with 68k emulation".. Warpup can't work with emulation either if no-one
ts> writes one for it. :)
Well, my point is that this would require HUGE work. Work which is already
done for WarpUP. For ppc.lib this is all ABSOLUTELY theoreticall, nobody would
write such a thing. Also, certain things would have to be changed in the
Kernel. Big things.
ts> It will be a long time before any PPC-only machines become available, if
ts> ever.. Escena board require 68k too, remember. Your point is quite
ts> irrelevant..
No. Think of what it means for "serious" applications when the huge
Contextswitch-Delays don't happen anymore...
Steffen Haeuser
Steffen Haeuser
tsu...@feandil.lnet.lut.fi wrote :
Hi!
>>Well, it was in talk at the start, that they gave software developpement
>>completely to H&P, dropping ppc.library... this would have been good, if they
ts> Why would they have done that?
Simply as:
- They would not have had to put more manpower in the early ppc.library and
their one (at that time two) coder(s) could concentrate on other things
- They could get a much more Amiga-Oriented solution for free then
Steffen Haeuser
Jyrki O Saarinen
> I remind you: StormC PPC was available much before SAS/C PPC and
> EGCS PPC !!!
Not true, gcc/egcs has been able to create ppc-elf code much longer.
Jyrki O Saarinen
> - They could get a much more Amiga-Oriented solution for free then
Which is more 'Amiga-oriented' whatever this means is a matter
of opinion, not a fact.
Jyrki O Saarinen
> jx> production quality compiler? (not to mention ceratain Mr. H.
> jx> yelling 'use WarpOS or die')
> Well, i did not say this. Also: You always had the attitude 'don't use
> StormC or die'. Who is the fanatic now ?
That's in your head. Find an article with dejanews where this
'attitude' shows.
Jyrki O Saarinen
> How did Haage and Partner's software split it?
How? This is a joke, right?
By confusing software developers and users.
Luckily there is powerpc.library wrapper for ppc.library and
alpha ppc.library wrapper for powerpc.library. So in theory
users can run both executable formats.
>> I have the feeling p5 never had any programmers on their
>> payroll, all work done 'out house'.
> Well, whether in house or not, they basically have one software
> engineer left. (And I use the term software engineer in it's
> loosest possible sense to describe the kludges he does.)
Kludges? Both kernels are as much kludges because the whole ppc
in the Amiga is huge kludge (like CGGX and Picasso96 too).
Jyrki O Saarinen
> jx> Escena boards will need a 68k in the CPU-slot for deliviring
> jx> interrupts from the Amiga motherboards to ppc in the Z3 slot.
> True. But this is no time-intensive stuff. The point is that OS-calls will
> speedup MUCH. The OS will - with exception of interrupts - run PPC Native.
The Os will run 68k emulated you meant?
Rocou
SH> k...@easynet.fr wrote :
SH> Hi!
ka>> SH> Hi!
>>>>> On my machine too. Only H&P's demos work.
>>fl>> Jupp, I have heard several other pepole say it crashes to.
ka>> SH> Well, this is - like i said in this thread already SEVERAL times -
ka>> the SH> problem of the LoadSeg-Patch.
ka>> [...]
ka>> SH> Well, if you install Terminator correctly - and don't activate ANY
ka>> SH> ppc.lib program before starting WarpUP - including CGX PPC and SH>
ka>> PPCInstall - then those problems don't appear...
ka>> OK but why H&P's demos work fine?
SH> They don't call LoadSeg inside PPC-Code ... only this causes problems.
SH> Inside PPC-Code ppc.library cannot be successfully opened, and as the
SH> LoadSeg-Patch of Ralp Schmidt calls ppc.library functions WITHOUT
SH> checking if the library COULD be opened, this causes the known problems.
Why do you call LoadSeg in your programs?
Rocou
SH> fa...@usa.net wrote :
SH> Hi!
>>fl>> I run the ppc.library all the time, there is no problem.
>>>Still it has not much to do anymore with the Amiga :)
>>>(And crashes for many people's systems...)
fa>> On my Amiga only WarpUp crashes.
SH> On mine only ppc.library.
Don't be ridiculous.
Do you think PPC card owners are stupid?
Everybody can check your assertion is wrong.
Casey R Williams
Hey guys!? I don't know if all of this is necessary. I've done my
best to keep on top of the goings-on regarding H&P, P5, VT etc.
My machine is a 4K with CSPPC 060/604e. I too have had better luck
getting H&P's WarpUp demos to run than ANY other WarpUP software.
Also, I am a Tornado3D fanatic, so I use the P5 PPC library. I have
never managed to terminate it, and haven't wished to move it. But,
I LOVE the performance of my machine when using T3D's PPC modes. (I
hardly think that something which can be made to work so well should
be dismissed as a kludge) Anyway, I would gladly embrace WarpUP,
should someone solve these compatibility problems. It gets much more
support from its creators, and appears to be favoured by "important"
developers (who know who they are ;). I wish laire had just swallowed
a little pride, as he could have been very helpful to H&P if he'd
wanted to be. Please concentrate on getting the two libs to coexist,
as there's going to be 2 of them to deal with until the next PPC
boards come out, at the earliest. As a user, I can't be bothered by
a -+10%, or even a -+20%, while both conflicting libraries are still
being developed for. It used to be a choice, now it's just a pain in
the ass.
etherius
Sam Jordan
> > I'm repeating again, that the confusion could have been *MUCH*
> > reduced, if P5 didn't fight WarpUp. We didn't want to confuse
>
> Sure.
>
> Now, if you were p5 wouldn't you fight because kernel number two makes
> selling your ppc cards harder. Can you see the point?
No.
Why should kernel number two make it harder? At that time, the quality
of WarpOS was *FAR* superior, so it effectively had made it easier
to sell their hardware.
> >> The possible future ppc cards (which I find quite unlikely, but nice
> >
> > Two projects are in progress, I can't see why it should be unlikely,
> > that at least one of these projects will appear.
>
> What is the 2nd one besides Escena?
The responsible persons/company doesn't want to release any information
yet.
> > Everyone can write a compiler for WarpUp, all informations are
> > available. It's not H&P's problem, when there are none. WarpUp
> > came out more than a year ago, time enough to adapt the compilers.
>
> A funny attitude IMHO.
>
> You create a new executable format while there would have been formats
H&P created a new executable format??? Please analyse ALL existing
WarpUp executables and please tell us the difference to the normal
AMIGA-OS executables.
I remember, some times ago, I almost went crazy about you, because you
simply didn't want to understand that the executable format
****************** DIDN'T CHANGE *********************
Can you maybe be so kind and tell me, why you simply don't want to
believe this? Why are you permanently saying, that H&P did create
a new executable format, when it is definitely wrong? WHY?? How
many times I will have to repeat that again? 10 times? 100 times?
I'm getting finally TIRED about that. If you definitely see, that
you are wrong, then you should not try to neglect that. Otherwise
the discussions are getting useless.
bye
--
Sam Jordan
Sam Jordan
> >I did not mean this. Try integrating a 68k Emulation in - current !!! -
> >ppc.library. You will have problems, as ppc.lib does in some aspects require a
> >68k Kernel "below". WarpUP does not require a 68k Kernel, it can run 100% PPC
>
> Argh. I'm 100% sure that the _current_ warpup doesn't work in PPC-only
> machine. It doesn't even have the emulation yet, how could it work?
My current WarpUp does, and there were not much additions to be done.
WarpOS was always very 68K-independent, while the ppc.library was
always based on a centralistic architecture with a lot of 68K control.
> Both systems can work with 68k emulation, if the necessary changes are
> made. API already supports this in both systems
You can easily write a library, which simply emulates 68K code for both
systems. But it is completely different to integrate kernel and emulator
into a complete solution, which allows to use PPC-only hardware. The
kernel
must be able to boot up completely independent from the 68K, it must run
the emulator as PPC task, allowing to multitasking between emulator and
native task. This is already working for WarpUp, we demonstrated this in
Koeln (by letting the AMIGA run completely on the PPC, just with the 68K
providing the interrupts). Instead Phase5 claimed to have two emulators
and
did never release any information about it.
And besides all that, it doesn't make much sense just to say, what
'theoretically would be possible'. Important is, what we will likely see
in not too distant future. And we will see a complete working solution
for PPC-only hardware, based on WarpOS. We will most likely never see
a similar solution based on the ppc.library. You can, of course, believe
something else, it's your choice to ignore the reality.
bye
--
Sam Jordan
Sam Jordan
> SJ> Basically I suppose, that we both had similar reasons: the
> SJ> quality was unsufficient at these days. I hope that this helps to
> SJ> understand why we decided to go for our own PPC solution, just as
> SJ> you decided to use a compiler, which provided the quality you
> SJ> needed.
> SJ>
>
> That comparison is very weak imho. You can run two programs compiled
The comparison *only* served to explain, why H&P decided to go for its
own system, because of quality problems, just as developers decide for
a compiler for quality reasons. Just as we didn't want to wait until
ppc.library gets usable, developers don't want to wait until a
compiler gets usable, but decide to use another option (compiler).
I started to discuss this, because several people still blame us for
the decision to go for our own solution with the argument 'H&P should
have waited until ppc.library gets usable'. And I want to make clear,
that no sane person is willing to wait in dimensions of years, when
options are available.
I didn't want to compare kernels and compilers and the text shouldn't
have implied that.
bye
--
Sam Jordan
Teemu Suikki
>> You create a new executable format while there would have been formats
>H&P created a new executable format??? Please analyse ALL existing
Hunk format is a new PPC executable format.
End of discussion..
Teemu Suikki
>WarpOS was always very 68K-independent, while the ppc.library was
>always based on a centralistic architecture with a lot of 68K control.
That's just because you didn't have access to 68k low-level stuff, that is,
68060.library.
>You can easily write a library, which simply emulates 68K code for both
>systems.
My point exactly.
>But it is completely different to integrate kernel and emulator
>into a complete solution, which allows to use PPC-only hardware. The
Too bad that PPC-only hardware doesn't exist yet, and probably won't exist
in the very near future.
Nicholai Benalal
SJ> I didn't want to compare kernels and compilers and the text
SJ> shouldn't have implied that.
Neither did I think so. What I mean is that you could affect the
market in a harmful way by introducing a new kernel but not by using
some other compiler. The person who introduces a new kernel should
think at least twice before doing it. That's why I couldn't find any
sence in your argument.
I've got the fullest respect for the work you've done and I'm not
questioning whether you know the difference between a kernel and a
compiler :-)
/Nicholai Benalal
Steffen Haeuser
tsu...@feandil.lnet.lut.fi wrote :
Hi!
ts> In article <3699C6F4...@spin.ch>, Sam Jordan wrote:
>>WarpOS was always very 68K-independent, while the ppc.library was
>>always based on a centralistic architecture with a lot of 68K control.
ts> That's just because you didn't have access to 68k low-level stuff, that is,
ts> 68060.library.
WarpUP was planned from the beginning to be independent of the 68k.
ts> My point exactly.
>>But it is completely different to integrate kernel and emulator
>>into a complete solution, which allows to use PPC-only hardware. The
ts> Too bad that PPC-only hardware doesn't exist yet, and probably won't exist
ts> in the very near future.
Not exists yet: Yes.
In the near future: No.
(Leaving the 68k as Interrupt Handler does not change my point...)
Steffen Haeuser
Steffen Haeuser
tsu...@feandil.lnet.lut.fi wrote :
Hi!
ts> In article <3699C367...@spin.ch>, Sam Jordan wrote:
>>> You create a new executable format while there would have been formats
>>H&P created a new executable format??? Please analyse ALL existing
ts> Hunk format is a new PPC executable format.
No. Look at the docs at H&P. It is exactly the same format like under
AmigaOS 68k. Only Object-files differ. Now end
of discussion.
Steffen Haeuser
Steffen Haeuser
etheriu...@labyrinth.net wrote :
Hi!
et> Hey guys!? I don't know if all of this is necessary. I've done my
et> best to keep on top of the goings-on regarding H&P, P5, VT etc.
et> My machine is a 4K with CSPPC 060/604e. I too have had better luck
et> getting H&P's WarpUp demos to run than ANY other WarpUP software.
Again: The problem is the Loadseg Patch installed by ppc.library to the
system. A lot of WarpUP programs use PPC Shared Libraries (for example
rtgmaster). Such software needs that the ppc.library LoadSeg Patch is NOT
installed. To achieve this:
- set env:powerpc/TERMINATOR correct
- do not install any ppc.library software which is activated at bootup
- if you still have problems, then this is no WarpUP problem, but one
with rtgmaster. Install MCPramlibpatch then.
Done.
Steffen Haeuser
Steffen Haeuser
k...@easynet.fr wrote :
Hi!
>fa>> On my Amiga only WarpUp crashes.
ka> SH> On mine only ppc.library.
ka> Don't be ridiculous.
ka> Do you think PPC card owners are stupid?
ka> Everybody can check your assertion is wrong.
But it is not. It claims "Cannot load ELF-File"...
Steffen Haeuser
Steffen Haeuser
k...@easynet.fr wrote :
Hi!
ka> SH> They don't call LoadSeg inside PPC-Code ... only this causes problems.
ka> SH> Inside PPC-Code ppc.library cannot be successfully opened, and as the
ka> SH> LoadSeg-Patch of Ralp Schmidt calls ppc.library functions WITHOUT
ka> SH> checking if the library COULD be opened, this causes the known problems.
ka> Why do you call LoadSeg in your programs?
It is a legal AmigaOS function. Assuming that a certain Lib never fails is NOT
legal, though.
Well, without it i cannot do Shared Libraries... and well, i won't scrap the
AmigaOS feature of Shared Libraries...
Steffen Haeuser
Steffen Haeuser
jxsa...@cs.Helsinki.FI wrote :
Hi!
jx> Steffen Haeuser <mag...@birdland.es.bawue.de> wrote:
>> jx> Escena boards will need a 68k in the CPU-slot for deliviring
>> jx> interrupts from the Amiga motherboards to ppc in the Z3 slot.
>> True. But this is no time-intensive stuff. The point is that OS-calls will
>> speedup MUCH. The OS will - with exception of interrupts - run PPC Native.
jx> The Os will run 68k emulated you meant?
At the start. I think from then surely a PPC Port - until it runs PPC Native -
will be following.
Steffen Haeuser
Steffen Haeuser
jxsa...@cs.Helsinki.FI wrote :
Hi!
jx> Steffen Haeuser <mag...@birdland.es.bawue.de> wrote:
>> I remind you: StormC PPC was available much before SAS/C PPC and
>> EGCS PPC !!!
jx> Not true, gcc/egcs has been able to create ppc-elf code much longer.
Did i EVER say GCC? I said EGCS (yes, i know EGCS is the Alpha/Beta of GCC...).
Steffen Haeuser