Archimedes Bah Humbug!
D.J.Davies
But I don't like the Archimedes computer
argargrhrhhghghhhh
Shock Horror
There is not much wrong with the machine only three things stop
me from bying one instead of the wonderfull Amiga I have at the
moment.......
1) The Machine is too unstable (probably due to badly written
software (see point two) ) I've been working with RiscOs 3.1
A3000's for a year and a half now.
2) The badly written comercial software, The envoirnment is
nice the chip is very fast (why no Co-Processors), of all the
packages that I've used only three impressed me.
3) The filosiphy (I can't spell to save my life) of the Arc is
that software is modular If your wordprocessor has'ent got a
table editor? then buy one. (What a pain in the Ass!)
Darren J Davies
I would buy an archimedes if there was no such thing as the
commodore Amiga...... 1280x512 in 262,144 colours, I love the
Amiga .........
/\/\/\/\/\/\/\/\/OO,
DJDa...@uk.ac.ex.cen
Paul Adams
: I don't mean to offend anyone here!
:
: But I don't like the Archimedes computer
:
Oh my God! It's one of them again !!!!
[things deleted]
: 2) The badly written comercial software, The envoirnment is
: nice the chip is very fast (why no Co-Processors), of all the
: packages that I've used only three impressed me.
:
We don't need Co-Processors to make the Archimedes go fast like the Amiga. The Archie can cope admirably with the one processor.
Paul Adams (CNS)
Philip R. Banks
> There is not much wrong with the machine only three things stop
> me from bying one instead of the wonderfull Amiga I have at the
> moment.......
>
> 1) The Machine is too unstable (probably due to badly written
> software (see point two) ) I've been working with RiscOs 3.1
> A3000's for a year and a half now.
It is? News to me. My machine does lock up now and then but mostly when I
have been working on some software or testing some badly written freeware
software. (And you kinda have to expect trouble with freeware, it one of the
reasons it is so cheap.)
But all the commercial software I use works like a charm.
> 2) The badly written comercial software, The envoirnment is
> nice the chip is very fast (why no Co-Processors), of all the
> packages that I've used only three impressed me.
Which three were these and why do they stand out? Also what is wrong with
the other software? It is all very well to make these claims but a little
actuall justification for their validity would be nice.
> 3) The filosiphy (I can't spell to save my life) of the Arc is
> that software is modular If your wordprocessor has'ent got a
> table editor? then buy one. (What a pain in the Ass!)
And you would prefer monolithic lumbering packages that chew HD space and do
everything? Well if that works for you no problems from me, but I kinda prefer
the modularity myself. For reasons I can go into if you like but they should be
obvious if you have done computer science.
> I would buy an archimedes if there was no such thing as the
> commodore Amiga...... 1280x512 in 262,144 colours, I love the
> Amiga .........
Thats cool, have fun with it. And I'll stick with what I find suits me best
and everyone is happy...'kay? Okay.
Philip
--
Philip R. Banks Syntax: mail < ban...@kosmos.wcc.govt.nz > @@@@@/|
Wesley suddenly underwent a gratutious total existence failure. "My God!@@@@/#|
" cried Dr. Crusher, "What's happened to Wesley?". "Who cares?" replied @@@/##|
Riker. "Agreed, #1," said Picard. "Mr. Data, take us out of here-Warp 9.@@/---|
Engage." The Enterprise shot towards the stars. --- Minimalist Trek @/ |
Jordan K. Hubbard
We don't need Co-Processors to make the Archimedes go fast like the
Amiga. The Archie can cope admirably with the one processor.
Um.. [clears throat] I'd kinda like an FPU, myself.. As soon as I can
find somebody who'll sell me one for my A540, I'm buying one. Any
suggestions? Everybody seems keener to offer upgrades for the A5000's
these days. So much for being at the forefront. Sob..
Jordan
--
(Jordan K. Hubbard) j...@violet.berkeley.edu, j...@al.org, j...@whisker.lotus.ie
David Alan Gilbert
>Paul Adams (CNS)
Is this still true? I mean wouldn't it be nice to have a graphics copro
which could speed drawing up - especially if it could do something
about the speed Draw renders its grid!!!!
Dave
--
-------------------------------------------------------------------------------
- David Alan Gilbert - gilb...@cs.man.ac.uk - G7FHJ@GB7CRG -
----------------(Don't let the bugs grind you down)----------------------------
Clive Jones
>Is this still true? I mean wouldn't it be nice to have a graphics copro
>which could speed drawing up - especially if it could do something
>about the speed Draw renders its grid!!!!
The FPA speeds that up a bit, as it happens. It would seem that
the grid plotting uses a lot of floating point.
--Clive.
(Disclaimer: I wouldn't believe a word of this if I were you...)
Gary J Palmer C.S.3
>
> Um.. [clears throat] I'd kinda like an FPU, myself.. As soon as I can
> find somebody who'll sell me one for my A540, I'm buying one. Any
> suggestions? Everybody seems keener to offer upgrades for the A5000's
> these days. So much for being at the forefront. Sob..
Hi Jordan
Just a quick word :
The FPA IS NOT A PROCESSOR!! Is is a maths accelerator chip (I know, I've
got one!), and while it can be classed as a co-pro, it does not come
under what the Amobea chappie was talking about.
I seem to remember some Amobeas allowing multiple 680x0 processors (I.E.
CPU's) . Whilst I wouldn't mind a multi-arm700 m/c for Arch-BSD, I think
*ONE* ARM3 will do fine (at the minute).
Honestly though, the only time I curse not having multi-processors
is during ray-tracing... which is by definition processor-heavy (in
any environment, apart from perhaps a Connection Machine!). Even then,
I think 320x256 in 20 mins is more than acceptable (try getting the amobea to
do that with with povray and even a simple scene!)
I honestly don't see what makes the amobeas so attractive... they have
a horrifying OS (or the Amiga 4000 I saw did), claims of high graphics
resolutions and BPP's, but which are all done in h/w using some sort
of cheating system which doesn't give True Colour (TM) and, most
importantly, they can't run !Axis or !Impress !!
Yours
Gary
--
JANET : gpa...@uk.ac.strath.cs | Internet : gpa...@cs.strath.ac.uk
Other Nets : gpalmer%cs.strath.ac.uk@ plus one of :-
BITNET: UKACRL UUCP: ukc.uucp | #include <std.disclaimer>
Oh sh*t, Windows NT has been released ... DEATH TO WIN NT
C.P. Brown
|> In article <JKH.93No...@whisker.lotus.ie>, j...@whisker.lotus.ie (Jordan K. Hubbard) writes:
|> >
|> > Um.. [clears throat] I'd kinda like an FPU, myself.. As soon as I can
|> > find somebody who'll sell me one for my A540, I'm buying one. Any
|> > suggestions? Everybody seems keener to offer upgrades for the A5000's
|> > these days. So much for being at the forefront. Sob..
|>
|> Hi Jordan
|>
|> Just a quick word :
|>
|> The FPA IS NOT A PROCESSOR!! Is is a maths accelerator chip (I know, I've
|> got one!), and while it can be classed as a co-pro, it does not come
|> under what the Amobea chappie was talking about.
|>
|> I seem to remember some Amobeas allowing multiple 680x0 processors (I.E.
|> CPU's) . Whilst I wouldn't mind a multi-arm700 m/c for Arch-BSD, I think
|> *ONE* ARM3 will do fine (at the minute).
I've never seen an Amiga with multiple 680x0's, apart from old 16 bit ones with a
32 bit CPU installed, keeping the old 68000 for backwards compatibility with bad
software. There was an april fools posting this year about a multiple 68040 and
060 Amiga, but that was just that, a joke.. There is a cute device called the
screamer from NewTek, which uworks in parallel with the video toaster and uses
several MIPS R4400's to speed up ray tracing...
|>
|> Honestly though, the only time I curse not having multi-processors
|> is during ray-tracing... which is by definition processor-heavy (in
|> any environment, apart from perhaps a Connection Machine!). Even then,
|> I think 320x256 in 20 mins is more than acceptable (try getting the amobea to
|> do that with with povray and even a simple scene!)
My A1200 takes about 4 minutes to do that resoloution in POVRAY. Of course, it's
faster using a real ray tracer like Imagine, for example :-)
In terms of processors/co processors, my machine has:
14MHz 68020 CPU
25MHz 68882 FPU
AGA Copper + Blitter
DMA controllers (disk IO, sound, etc)
oh, and a 6502 to read the keyboard :-)
|>
|> I honestly don't see what makes the amobeas so attractive... they have
|> a horrifying OS (or the Amiga 4000 I saw did), claims of high graphics
|> resolutions and BPP's, but which are all done in h/w using some sort
|> of cheating system which doesn't give True Colour (TM) and, most
|> importantly, they can't run !Axis or !Impress !!
Maybe you should try actually USING AmigaDOS, instead of just flaming it. It is
one of the most powerful single user OS'es in existance, cannot be matched for
multitasking, has lots of nice stuff such as datatypes, screens etc, and a really
powerful shell language (more powerful than a lot of unix stuff in a lot of
ways), that is also easy to use. As for graphics, 1400*592, 800*600 and 784*1100
in 262144 colours, or 256 REGISTERED colours from a 24 bit palette seem perfectly
adequate for base model graphics. If you want very high res 24 bit graphics, then
buy a GVP spectrum EGS card (1600*1280 in 256 colours, lower resoloutions in 15
and 24 bit), or some such. There is a lot of choice in this area. The AAA chipset
will be available sometime in 94, which gives 1280*1024 24 bit non-interlaced,
programable pixel clock, 8 channel 100KHz 16 bit sound, and lots of other nice
features. It's 64 bit and can support up to 16 megabytes of CHIP RAM, which can
be VRAM, DRAM, or a mixture of both.
--
Chris Brown
-------------------------------------------------------------------------------
Amiga 1200, 6MB, 80 Meg harddisk, 25MHz 68882, 262144 colours on screen
-------------------------------------------------------------------------------
Finger me at cpb...@hermes.cam.ac.uk for my PGP public key
Jordan K. Hubbard
The FPA IS NOT A PROCESSOR!! Is is a maths accelerator chip (I know, I've
got one!), and while it can be classed as a co-pro, it does not come
under what the Amobea chappie was talking about.
Frankly, I don't think the `Amobea' chap was talking about that
either. We're just talking about a plain 'old FPU, as is commonly
defined as a 68881/68882 in Moto land, a R2010 in MIPS land, an i387
in Intel land, etc. Real multiprocessing (with a bonafide fully
functional CPU as opposed to some semi-intelligent hardware) is not,
to my knowledge, the providence of ANY currently available
off-the-shelf "consumer electronics catagory" hardware. I want an
FPU so my math calculations go faster, that's all. True multiprocessing
can wait for tomorrow's micros..
I seem to remember some Amobeas allowing multiple 680x0 processors (I.E.
CPU's) . Whilst I wouldn't mind a multi-arm700 m/c for Arch-BSD, I think
Not true. There have been upgrades that *replaced or disabled*
existing processors, but none that allowed execution in parallel. I'm
not counting some of the more recent insane Amiga NewTek ("screamer")
hardware here, since that's out of the reach of most mortals.
Honestly though, the only time I curse not having multi-processors
is during ray-tracing... which is by definition processor-heavy (in
any environment, apart from perhaps a Connection Machine!). Even then,
Actually, a good FPU will give you excellent performance boosts for
things like ray tracing, single integer processor or no. To be
honest, I don't expect a lot out of ARM's FPU offering, given the
price range they're offering it at, but some accelleration is a hell
of a lot better than none. I wish someone with an FPU would post some
ray-tracing oriented benchmarks..
If you want to ray trace your brains out, you buy something like an
ALPHA with very nice FPU performance indeed. Otherwise you upgrade
as best you can and hope for the best.
I think 320x256 in 20 mins is more than acceptable (try getting the amobea to
do that with with povray and even a simple scene!)
My Amiga A2500 with 68030/68882 @ 25Mhz beats the pants off the A540
using POV-Ray 1.0. Believe me, those are not the results I *wanted*
to see, I was rather hoping for more out of the Arch, but the lack of
an FPU apparently kills it pretty good (yes, we optimized the code).
I think the A4000, which has an 68040 with inbuilt (faster) FPU would
do even better than this. The Arch is pretty damn fast, but not in
all catagories, unfortunately.
Don't believe me? Ok: Try an 800x600 scene with anti-aliasing of 0.1
of something like the dragon (common POVRay object). On my A540, this
took over 15 hours. On the Amiga 2500 it took 7.5 hours. On the
486/66 DX2 it took 3.2 hours, and on a (borrowed :-() HP 735 it took
45 minutes!
We also got someone to run the same scene on their ALPHA AXPnnn
(product number unknown) and they said 21 minutes, but this is still
unverified.
I honestly don't see what makes the amobeas so attractive... they have
a horrifying OS (or the Amiga 4000 I saw did), claims of high graphics
I think you have either never really seen a horrifying OS, or you
(more likely) simply did not look very closely. Look again, and this
time READ THE PROGRAMMER'S DOCUMENTATION before you comment on
something like this. Going up to a machine in a shop, frowning at it
for 10 minutes, and then saying "the OS on this thing *sucks*!" is
hardly what one would consider the most impartial route to an informed
opinion.
resolutions and BPP's, but which are all done in h/w using some sort
of cheating system which doesn't give True Colour (TM) and, most
None of the currently available machines have True Color (TM) as a
standard feature. The Amiga, at least, left its architecture open
enough that it could be added later, and for $400 I did just that when
I went out and bought a Retina(tm) card for it, something which gives
perfectly servicable 800x600x24, 1024x768x16 or 1280x1024x8 graphics
resolutions. This is called _expandability_ and is a Very Significant
factor in considering the viability of a machine.
If I could do such a thing for my A540, I would have long ago but,
sadly, a G16 card was the best I could obtain due to limitations in
the Archimedes design. Don't get me wrong - it's a nice card and it
increases the usability of the machine tenfold, but it doesn't come
close to matching something like the Retina's (or the Piccolo, or the
Merlin, or the Impact 24, or the.. you get the point) specs.
Sometimes designers screw the pooch, even on your Favorite Machine(tm),
and it's just folly to pretend otherwise with bluster and finger pointing
in someone else's direction. Better to find the shortcomings and then
regularly attend the developer's conferences, where you someday may be
able to corner the V.P. of engineering, point your finger at him and
say, in a very loud voice, "J' Accuse'!"
N W H Mailer
school, from an old A310 right up to an A5000. We pushed the machines, went to
all the shows, met all the gurus etc. I wrote my A Level computing project on it,
and I have to conclude that, although in some ways it is a wonderful machine, in
other ways, one way in particular, it absolutely SUCKS!
This one way in particular is the fact that its OS, right from Arthur (now THERE
was an operating system from hell) was based far too closely on the old BBC
micro's. This was ridiculous, and lead to a marginalising of the GUI and an
absolutely terrible implementation of BASIC (anyone who has tried as much as
opening a window with BASIC V will know what I'm saying). Also, the whole rigid
graphics mode structure was absolutely antiquated, and there is absolutely no
hardware to help with scrolling etc...
I have used an Amiga at home since they were introduced, and I can safely claim
that there are things about that machine which are just as revolutionary as RISC
- the DMA co-processors, the multi-mode graphics SIMULTANEOUSLY (pulling down the
WorkBench screen to reveal a HAM pic beneath causes a gasp among many Arch
owners). Also, I must say that AmigaDOS 3.00 (on the Amiga 1200 and 4000) beats
the living daylights out of RISCOS. And yes, I have studied the Arch's
programming reference manuals and used RISCOS quite intensely. There was one
thing about the Arch's GUI which I liked - dragging icons from an APP to save
etc... This is something which the Amiga has now implemented too.
BUT .. I am not saying that the Amiga is better than the Arch or visaversa. There
is some superb software on the Arch (Artworks, Impression etc..) which prove how
good the machine is. What I am saying is that both machines were and are
mould-breaking, and in their own ways revolutionary. Some things one does better,
other things the other. What I can say is that original, mould breaking and
quirky systems like the Arch and the Amiga beat the bland, standardized and
patronising systems: the PCs and the Macintosh. Both are examples of dull,
generic hardware which is supported to hell and back. If the Amiga and Arch had
that level of support and money, the computer industry would be a healthier
place. Come on guys, stop bashing the other innovators - try bashing the real
threat to computer creativity instead!!
My perfect system: An Acorn A5000, an Amiga 4000 and a file converter!
Jon Ribbens
eng3...@sun.leeds.ac.uk (N W H Mailer) writes:
>I CANNOT be called biased or making uninformed conclusions. I used Archimedes at
>school, from an old A310 right up to an A5000. We pushed the machines, went to
>all the shows, met all the gurus etc. I wrote my A Level computing project on it,
>and I have to conclude that, although in some ways it is a wonderful machine, in
>other ways, one way in particular, it absolutely SUCKS!
Gee, that's harsh...
>This one way in particular is the fact that its OS, right from Arthur (now THERE
>was an operating system from hell) was based far too closely on the old BBC
>micro's. This was ridiculous, and lead to a marginalising of the GUI and an
Umm... Gee, that sounds like a biased and uninformed conclusion if ever I
heard one ;)... How exactly does RISC OS having a few calls similar to the
BBC OS make it suck?
>absolutely terrible implementation of BASIC (anyone who has tried as much as
Gee, and I always thought Acorn BASIC was one of the best implementations in
the world... Now it turns out it's absolutely terrible. You learn something
new every day.
>opening a window with BASIC V will know what I'm saying). Also, the whole rigid
Dang, and I always used to use SYS "Wimp_OpenWindow",,q%, and thought it was
easy. Shows how wrong you can be, huh? BTW, anybody here tried opening a
window on an Amiga? <g>
>graphics mode structure was absolutely antiquated, and there is absolutely no
>hardware to help with scrolling etc...
<g> You're speaking from the standpoint of an Amiga owner, the *only* home
machine I know of that has hardware to help with scrolling... The ARM gets
around this by being horribly fast.
>I have used an Amiga at home since they were introduced, and I can safely claim
>that there are things about that machine which are just as revolutionary as RISC
>- the DMA co-processors, the multi-mode graphics SIMULTANEOUSLY (pulling down the
>WorkBench screen to reveal a HAM pic beneath causes a gasp among many Arch
<g> Hum, personally, my biased and uninformed view is that the screen system
on the Amiga's desktop is vile... But there you go. Of course, the screens
*are* useful if you're doing parallax scrolling in a game... But Amiga's aren't
designed just for games, are they? Nooooo.... ;)
>owners). Also, I must say that AmigaDOS 3.00 (on the Amiga 1200 and 4000) beats
>the living daylights out of RISCOS. And yes, I have studied the Arch's
<g> Ah well, I used AmigaDOS, umm, 2.00 I think, and it was utter shite (just
my biased and uninformed view ;) ). Maybe it's suddenly changed completely in
version 3, or maybe you just have to like the horrible colours ;).
>programming reference manuals and used RISCOS quite intensely. There was one
>thing about the Arch's GUI which I liked - dragging icons from an APP to save
>etc... This is something which the Amiga has now implemented too.
Wahay! Hmm... It's odd that, dragging *from* the app *to* the filer window
seems utterly sensible and yet things like OS/2 and Windows *still* don't
have it.
>BUT .. I am not saying that the Amiga is better than the Arch or visaversa. There
Doesn't sound like it from your first paragraph ;). The Amiga is not better
than the Arc but the Arc sucks?
>is some superb software on the Arch (Artworks, Impression etc..) which prove how
>good the machine is. What I am saying is that both machines were and are
>mould-breaking, and in their own ways revolutionary. Some things one does better,
>other things the other. What I can say is that original, mould breaking and
>quirky systems like the Arch and the Amiga beat the bland, standardized and
>patronising systems: the PCs and the Macintosh. Both are examples of dull,
>generic hardware which is supported to hell and back. If the Amiga and Arch had
>that level of support and money, the computer industry would be a healthier
>place. Come on guys, stop bashing the other innovators - try bashing the real
>threat to computer creativity instead!!
<g> Yeah, let's all kill B... G.... shhh... you know who
>My perfect system: An Acorn A5000, an Amiga 4000 and a file converter!
Hehe... Surely an A5000 with an Amiga plug-in card? Only one keyboard then,
easier to use...
--
// Jon Ribbens // Email: es...@csv.warwick.ac.uk or j.ri...@warwick.ac.uk //
// Term time: E09 Draycott House, Rootes Residences, University of Warwick, //
// Coventry, CV4 7AL // Home: 59 Upper Belmont Road, Chesham, Bucks HP5 2DD //
Adam Curtin
>In article <1993Nov27....@gps.leeds.ac.uk>,
> eng3...@sun.leeds.ac.uk (N W H Mailer) writes:
>>This one way in particular is the fact that its OS, right from Arthur (now THERE
>>was an operating system from hell) was based far too closely on the old BBC
>>micro's. This was ridiculous, and lead to a marginalising of the GUI and an
>Umm... Gee, that sounds like a biased and uninformed conclusion if ever I
>heard one ;)... How exactly does RISC OS having a few calls similar to the
>BBC OS make it suck?
You can't tell whether it was biased and uninformed from that statement - it
sounds like Jon Ribbens is in a good position to judge. It was an unjustified
statement, but not an unjustifiable one.
I dunno about "GUI marginalisation", but I've always thought the
ten-character filename and seventy-seven objects per directory limit were
particularly sucky.
>>absolutely terrible implementation of BASIC (anyone who has tried as much as
>>opening a window with BASIC V will know what I'm saying). Also, the whole rigid
>Dang, and I always used to use SYS "Wimp_OpenWindow",,q%, and thought it was
>easy. Shows how wrong you can be, huh? BTW, anybody here tried opening a
>window on an Amiga? <g>
It *is* a pain in the bum programming wimp applications from BASIC - only
folk who haven't used a language with structures could possibly think
otherwise. Using the Wimp from BASIC is essentially just interpreted
assembler programming, and I think that's what the original poster was
getting at ... there's no built-in functions or procedure library for Wimp
programming. I've never had the privilege of using an Amiga though, so I
don't know how it compares.
>>My perfect system: An Acorn A5000, an Amiga 4000 and a file converter!
You have very low ambitions!
A.
---
/home/adam/.sginature: No such file or directory
Simon Glass
> Is this still true? I mean wouldn't it be nice to have a graphics copro
> which could speed drawing up - especially if it could do something
> about the speed Draw renders its grid!!!!
That's just Draw. It could be sped up around 5x by altering the code a bit,
without using FP.
--
Simon
Sig line trunca
Jim Noble
>My Amiga A2500 with 68030/68882 @ 25Mhz beats the pants off the A540
>using POV-Ray 1.0. Believe me, those are not the results I *wanted*
>to see, I was rather hoping for more out of the Arch, but the lack of
>an FPU apparently kills it pretty good (yes, we optimized the code).
Comparing a machine *with* FPU to one *without* FPU is a bit of a silly
thing to do. What are talking about anyway? 'lack of FPU' What is the FPA10
then?
>Don't believe me? Ok: Try an 800x600 scene with anti-aliasing of 0.1
>of something like the dragon (common POVRay object). On my A540, this
>took over 15 hours. On the Amiga 2500 it took 7.5 hours. On the
>486/66 DX2 it took 3.2 hours, and on a (borrowed :-() HP 735 it took
>45 minutes!
Ok. The FPA10 increases Floating point maths by a factor of 2-5. Take 3 as
an average. This should mean that on an A540, the ray tracing will take 5
hours. Faster than your amiga :-)
Jim
--
Jim Noble. 3rd year Computer Systems Engineering. University of Kent.
Address: 45 Sydenham Street, Whitstable, Kent CT5 1HN.
What I just said is probably wrong.
Heck! Everybody knows I only open my mouth to change feet. :-)
R. Brown
|> In article <1993Nov27....@gps.leeds.ac.uk>,
|> eng3...@sun.leeds.ac.uk (N W H Mailer) writes:
[...]
|> >My perfect system: An Acorn A5000, an Amiga 4000 and a file converter!
|>
|> Hehe... Surely an A5000 with an Amiga plug-in card? Only one keyboard then,
|> easier to use...
Just thought I'd make the pint: The Amiga's the emulation specialist. Give us an
ARM CPU for Emplant (Or let the guys at UU have a go at emulating it. They seem
to think they can get an almost full speed 486 emulation out of a 68040...<grin>)
and a set of RISCOS ROMS, and we'll probably manage. Now, what would the Arch
need for an Amiga emulation?
Seriously though, I think the point is well made. If I couldn't have an Amiga, my
second choice of computer would be an Arch. Why would ANYONE want to use Windoze?
:-)
|>
|> --
|>
|> // Jon Ribbens // Email: es...@csv.warwick.ac.uk or j.ri...@warwick.ac.uk //
|> // Term time: E09 Draycott House, Rootes Residences, University of Warwick, //
|> // Coventry, CV4 7AL // Home: 59 Upper Belmont Road, Chesham, Bucks HP5 2DD //
Robert Brown.
A G Jackson
|> Ok. The FPA10 increases Floating point maths by a factor of 2-5. Take 3 as
|> an average. This should mean that on an A540, the ray tracing will take 5
|> hours. Faster than your amiga :-)
How about the new faster FPA for the 33Mhz ARM3 machines? Is this still due
out some time, and if so how much will it cost and how much will it increase
Floating point maths by?
Adrian
+--------------------------------------------------------------------------+
| Adrian Jackson | a.g.j...@durham.ac.uk | 27 North Bailey, DURHAM |
+--------------------------------------------------------------------------+
| The Archchancellor's most important job, as the Bursar saw it, was to |
| sign things, preferably, from the Bursar's point of view, without |
| reading them first. |
| -- (Terry Pratchett, Moving Pictures) |
+--------------------------------------------------------------------------+
Jim Noble
>How about the new faster FPA for the 33Mhz ARM3 machines? Is this still due
>out some time, and if so how much will it cost and how much will it increase
>Floating point maths by?
It will still increase floating point maths bya factor of 2-5 :*) It's just
that floating point maths would be faster at 33MHz than at 25MHz.
The 33MHz FPA might be available early next year. Who knows, this is Acorn
we are talking about remember :*)
Torben AEgidius Mogensen
>The 33MHz FPA might be available early next year. Who knows, this is Acorn
>we are talking about remember :*)
No. It is whoever is producing the 33MHz FPA. I'm sure we will see
third-party upgrades within weeks after the chips become generally
available.
Torben Mogensen (tor...@diku.dk)
Jim Noble
>No. It is whoever is producing the 33MHz FPA. I'm sure we will see
>third-party upgrades within weeks after the chips become generally
>available.
Yes, but it is Acorn who commissioned it, and Acorn who know the details.
Jim
--
Jim Noble. 3rd year Computer Systems Engineering. University of Kent.
What I just said is probably wrong.
Kyle Dawkins
> [Stuff deleted]
>Seriously though, I think the point is well made. If I couldn't have an Amiga, my
>second choice of computer would be an Arch. Why would ANYONE want to use Windoze?
Well, let's face some facts. Windows is by no means the be-all end-all
of PC operating systems. I think you pose the question in a different
light... OS/2 is incredibly cheap and in terms of multi-tasking
is at least as capable as any Amiga you care to name.
And that's not the only pre-emptive scheduling OS for the
PC... linux is free and extremely good, there's always been minix,
now there's 386BSD, all free. There's WNT (Cough cough splutter
splutter) if you're congenitally insane. Desqview has always
been an appealing alternative to DOS/Windows. And cheap too.
But to top it all off, the PC is the only machine now available
that runs NeXTStep, which is the nicest OS I have ever laid eyes
on, from the inside and the outside. Don't rubbish PCs because
of their varied history. There is some remarkable software
available for them.
I agree with you... why WOULD anyone want to use Windows? But
I don't think it's valid to judge PCs because of it.
#serious mode off
I love both the Arc and the Amiga, too, so don't worry. I own
an Amiga, but only because you can't get Arcs here in Canada!!!
What I want to know is, is there any move from Acorn's point
of view to encourage people to port other OS's to the Arc?
I mean, I think a high-end Arc could do a pretty mean
port of NeXTStep or OS/2. I suppose they wish to keep
RO3 and Riscix and not let anyone else muscle in.
Oh well.
Cheerio!
Kyle
John Robinson
>In article <2d5jh4$s...@dunlop.cs.strath.ac.uk>, gpa...@cs.strath.ac.uk (Gary J Palmer C.S.3) writes:
>|> In article <JKH.93No...@whisker.lotus.ie>, j...@whisker.lotus.ie (Jordan K. Hubbard) writes:
>|> >
>|> > Um.. [clears throat] I'd kinda like an FPU, myself.. As soon as I can
>|> > find somebody who'll sell me one for my A540, I'm buying one. Any
>|> > suggestions? Everybody seems keener to offer upgrades for the A5000's
>|> > these days. So much for being at the forefront. Sob..
Forefront? What forefront? Acorn have dropped the A540, and in fact did so ages
agoo. Sorry.
>|> The FPA IS NOT A PROCESSOR!! Is is a maths accelerator chip (I know, I've
>|> got one!), and while it can be classed as a co-pro, it does not come
>|> under what the Amobea chappie was talking about.
It *is* a processor. It's just that it only does FP instructions.
>|> I seem to remember some Amobeas allowing multiple 680x0 processors (I.E.
>|> CPU's) . Whilst I wouldn't mind a multi-arm700 m/c for Arch-BSD, I think
>|> *ONE* ARM3 will do fine (at the minute).
Well, there's no reason the next Arc couldn't have multiprocessor support; I
don't think the RISC OS programming interface would disallow it or anything.
>I've never seen an Amiga with multiple 680x0's, apart from old 16 bit ones with a
>32 bit CPU installed, keeping the old 68000 for backwards compatibility with bad
>software. There was an april fools posting this year about a multiple 68040 and
>060 Amiga, but that was just that, a joke.. There is a cute device called the
>screamer from NewTek, which uworks in parallel with the video toaster and uses
>several MIPS R4400's to speed up ray tracing...
Yeah, and I'll bet it's not cheap, either...
>|> I honestly don't see what makes the amobeas so attractive... they have
>|> a horrifying OS (or the Amiga 4000 I saw did), claims of high graphics
>|> resolutions and BPP's, but which are all done in h/w using some sort
>|> of cheating system which doesn't give True Colour (TM) and, most
>|> importantly, they can't run !Axis or !Impress !!
Well, the HAM modes are a bit of a cheat, but what do you usually want to
display using all your millions of colours? Scanned photographic pictures and
bits of artwork you've drawn yourself. Graduated fills are trivially easy, and
you get a basic palette to draw your desktop in too. So I do object to the way
it's hailed as 262,144 colours when you couldn't actually display a random
screen with all those colours on it, but when you come down to using it on a
desktop, it does everything yer 256 colour mode does, and a bit more, so it can
be useful.
>Maybe you should try actually USING AmigaDOS, instead of just flaming it. It is
>one of the most powerful single user OS'es in existance, cannot be matched for
>multitasking, has lots of nice stuff such as datatypes, screens etc, and a really
>powerful shell language (more powerful than a lot of unix stuff in a lot of
>ways), that is also easy to use. As for graphics, 1400*592, 800*600 and 784*1100
>in 262144 colours, or 256 REGISTERED colours from a 24 bit palette seem perfectly
>adequate for base model graphics. If you want very high res 24 bit graphics, then
>buy a GVP spectrum EGS card (1600*1280 in 256 colours, lower resoloutions in 15
>and 24 bit), or some such. There is a lot of choice in this area. The AAA chipset
>will be available sometime in 94, which gives 1280*1024 24 bit non-interlaced,
>programable pixel clock, 8 channel 100KHz 16 bit sound, and lots of other nice
>features. It's 64 bit and can support up to 16 megabytes of CHIP RAM, which can
>be VRAM, DRAM, or a mixture of both.
Well, since we're talking about future machines now, the next Arc will have a
VIDC20 in it. As I understand it, it can support 4Mb of VRAM, therefore
1600x1200x16bpp and 1280x1024x24bpp. The VIDC20 is available NOW and will be
fitted in new Acorn machines sometime next year. (OK, so it hasn't officially
been announced, but they're not stupid enough not to, iyswim.)
16Mb of RAM. Ho hum, not much really; we've had the option since 1989 or so
(A540); but no-one took it up 'cos most of the time you don't need it.
John.
--
Disclaimer: I know only that I know nothing
Clive Jones
>ARM CPU for Emplant (Or let the guys at UU have a go at emulating it. They seem
>to think they can get an almost full speed 486 emulation out of a 68040...<grin>)
>and a set of RISCOS ROMS, and we'll probably manage. Now, what would the Arch
>need for an Amiga emulation?
It's more plausible that an Amiga could be emulated on an Arc than
vice-versa, I feel. Blitting is easier to do in software on an ARM
than ARMcode is to do in a blitter. (-8
I don't believe that a 486 could be emulated at more than about a
tenth of full speed on a 68040.
--Clive.
A.W. Oakden
> of PC operating systems. I think you pose the question in a different
> light... OS/2 is incredibly cheap and in terms of multi-tasking
> is at least as capable as any Amiga you care to name.
> And that's not the only pre-emptive scheduling OS for the
> PC... linux is free and extremely good, there's always been minix,
> now there's 386BSD, all free. There's WNT (Cough cough splutter
> splutter) if you're congenitally insane. Desqview has always
> been an appealing alternative to DOS/Windows. And cheap too.
> But to top it all off, the PC is the only machine now available
> that runs NeXTStep, which is the nicest OS I have ever laid eyes
> on, from the inside and the outside. Don't rubbish PCs because
> of their varied history. There is some remarkable software
> available for them.
Well lets face some facts;
OS2 requires a minimum of 8Meg Ram to work and a lot more to work
properly!
OS2 takes approximately 40Meg of your hard drive to work at all!
Not exactly compact is it?
Tonyt Oakden
A G Jackson
|> In article <1993Nov29.2...@infodev.cam.ac.uk> Robert Brown <rb1...@hermes.cam.ac.uk> writes:
|> > [Stuff deleted]
|> >Seriously though, I think the point is well made. If I couldn't have an Amiga, my
|> >second choice of computer would be an Arch. Why would ANYONE want to use Windoze?
|> >Robert Brown.
|>
|> Well, let's face some facts. Windows is by no means the be-all end-all
|> of PC operating systems. I think you pose the question in a different
|> light... OS/2 is incredibly cheap and in terms of multi-tasking
|> is at least as capable as any Amiga you care to name.
|> And that's not the only pre-emptive scheduling OS for the
|> PC... linux is free and extremely good, there's always been minix,
|> now there's 386BSD, all free. There's WNT (Cough cough splutter
|> splutter) if you're congenitally insane. Desqview has always
|> been an appealing alternative to DOS/Windows. And cheap too.
|> But to top it all off, the PC is the only machine now available
|> that runs NeXTStep, which is the nicest OS I have ever laid eyes
|> on, from the inside and the outside. Don't rubbish PCs because
|> of their varied history. There is some remarkable software
|> available for them.
Yes. PCs run NeXTStep. Certainly. Provided they're a top of the range 486. With
vast amounts of hard disc space and memory. And a CD-ROM drive to install it
from. And the right architecture. Let's face it, you won't get much change out
of three grand. Then you've got to buy the OS...
Speaking personally, I'm not that impressed with NeXTStep. I know it's meant to
be the greatest OS ever. So will somebody explain to me why the first time I saw
it being demonstrated by a NeXT fan, it froze up the machine and died because it
didn't have enough free disc space? I mean, obviously this is far superior to the
RISC OS method of printing an error box containing the cryptic words "Disc Full".
|> What I want to know is, is there any move from Acorn's point
|> of view to encourage people to port other OS's to the Arc?
|> I mean, I think a high-end Arc could do a pretty mean
|> port of NeXTStep or OS/2. I suppose they wish to keep
|> RO3 and Riscix and not let anyone else muscle in.
Current high-end Arcs probably couldn't take NeXTStep. Current high-end Arcs
running OS/2 would be reduced to low-end Arcs in minutes. Myself I'd prefer an
updated RISC OS. It's one of the fastest and nicest operating systems around. Why
change? Alright, perhaps it needs pre-emptive multi-tasking, but I'm still not
convinced.
Adrian
+--------------------------------------------------------------------------+
| Adrian Jackson | a.g.j...@durham.ac.uk | 27 North Bailey, DURHAM |
+--------------------------------------------------------------------------+
| Dhblah sidled closer. This was not hard. Dhblah sidled everywhere. |
| *Crabs* thought he walked sideways. |
| -- (Terry Pratchett, Small Gods) |
+--------------------------------------------------------------------------+
N W H Mailer
Not true, I'm afraid! Take a look at Emplant's PC emulator for the Amiga and
compare it to any PC emulator available for the Archimedes. It MAY be possible
for the ARM to "blit", but because it's only one chip, something else would have
to be sacrificed. Unfortunately, three into one simply DO NOT go (well, outside
the metaphystical world, anyway). Proof of the pudding, the Arc has already been
emulated at the prototype stage, but the Amiga never has on the Arc.
Oliver Betts
R. Brown <rb1...@hermes.cam.ac.uk> wrote:
>Seriously though, I think the point is well made. If I couldn't have an
>Amiga, my
>second choice of computer would be an Arch. Why would ANYONE want to use
>Windoze?
>:-)
I propose a little experiment. I'll lock you in a room with a PC
running MS Windows and an Archimedes, but I'll take away the mice from
both machines. I'll come back after a week, and see which one you're
using.
Ol
--
Trained in Cambridge as an urban attack cyclist.
Torben AEgidius Mogensen
Maybe this is because Arc owners are not interested ;-).
It would be interesting to hear how fast and complete this prototype
emulation is. Is it only ARM code, or is the full Arc hardware
simulated?
Anyway, Clive said "emulated". With dynamic translation you can get
better speed. Running 486 code on a 68040 at anything better than 20%
speed (at similar clock speeds) can't be done with pure emulation.
Note that this is not a criticism of dynamic translation (or other
kinds of object code translation).
If you restrict yourself to Windows programs, you can get good results
with a combination of emulation of application code and rewriting the
Windows system calls in native code (like WABI).
But back to the issue of what is easiest: Arc simulation on an Amiga
or Amiga simulation on an Arc, I tend to agree with Clive that the
latter is easiest.
Simulation can be broken into three parts:
1) Instruction set simulation.
2) OS simulation.
3) Hardware simulation.
With good hardware simulation, OS simulation is reduced to emulating
the instructions and hardware (as the current Acorn PC emulator does).
This is, however, a slow method. A faster metod is to rewrite (parts
of) the OS in native code, and catch the calls to the OS. This is the
approach in WABI and for running Macintosh programs on the PowerPC.
This can dramatically speed up the part of execution that happens in
the OS, but has problems with user code using OS code through other
means than the documented calls or user code accessing hardware
directly. You can, however, combine these methods, so (some)
documented calls get caught and other accesses get emulated the hard
way.
Hardware simulation covers several points: writing to and reading from
devices and memory lay-out for graphics.
On PCs, graphics are handled by graphics cards with local memory,
which can be read and modified only through special device calls.
Hence you just have to simulate these calls.
With machines that (like both Arc and Amiga) use memory mapped
graphics, user code can change the appearance of the screen by
modifying memory without the need to write directly to a device (as
the device reads the memory by itself). The display hardware on Amiga
and Archimedes requires different memory layout for similar pictures,
so a simulator must translate between the formats on every memory
access to the relevant parts of memory. Testing for this situation at
every memory access is costly, so an efficient simulation must use
tricks to get around this.
One such trick is to keep two copies of screen memory: one in the
emulated machines format and one in the emulating machines format.
Reading from memory can then happen normally, as the emulated machines
copy is used. By marking this area read-only in the MMU, writes to
this part of memory will be trapped. The trap handler can complete the
writing and also convert the write to the emulating machines format
and thus update its own copy. This method is good if user code writing
is uncommon. Screen updating through OS calls can more directly update
both copies, so the overhead is less here.
A variation of this trick is to let the emulated machine freely modify
its own copy, and then at regular intervals (e.g. 5 times per second)
copy the entire screen memory to the emulating machines copy,
converting the format along the way. This method is good if user code
screen memory update is frequent. By making the rate of copy
selective, smoothness versus speed can be adjusted.
Note that these problems occur both when an Arc is emulating an Amiga
and vice-versa.
Emulating the blitter is fairly easy compared to emulating user code
access to screen memory, as the blitter code easily can update both
copies.
Instruction set simulation is the easiest part. Any good programmer
can over a weekend make a simulation of an instruction set. IT takes a
bit more effort to make it fast.
A traditional emulation simulates each instruction on its own. A naive
approach will compare the op-code field of the instruction with the
various possibilities, and then decode the operand fields and perform
the relevant action. This is, however, very slow. An average of about
100 instructions is needed to emulate a single instruction. Various
methods can improve on this.
If the instruction word is small, e.g. 8 or 16 bit a jump table can be
used. Each individual instruction has a bit of code that simulates
exactly this instruction (even down to specific registers). This code
is accessed by using the instruction word as index in a jump table.
For an 8-bit processor, this might even be the most compact emulator.
For a 16-bit instruction word (like 680x0), it requires somewhat more
memory. Let's say that an 680x0 instruction requires an average of 5
ARM instructions to emulate, plus one word in the jump table. This
requires 1.5MB of memory, so it is out of the question on small
machines, but perfectly feasible on 4MB Arcs. The Amiga can't use a
similar scheme to emulate ARM instructions as a 32-bit index is silly.
It can, however, use a hybrid scheme, where the first few bits of the
opcode are used as an index, and the rest are decoded.
Dynamic translation means that the emulator translates each
instruction to native code when it is emulated, wnd then executes this
code. The advantage is that it can keep the translated instructions,
and reuse the translation next time the instruction is executed.
Essentially you have a table with an entry for every word in the code,
mapping the instruction to a location where the translated code can be
found. Initially these all point to the generic emulator code, but as
the emulation progresses, some are changed to point to specialized
code. Whenever a meory location is written to, the corresponding
entry in the table is reset, so self-modifying code is handled
properly. Advanced versions of this scheme translates a basic block at
a time. This reduces overhead, but it is more complicated to maintain
correctness of self-modifying code.
Object code translation tries to translate the full program before
executing it. It must still maintain a table mapping original
addresses to translated code, as it otherwise can't cope with indirect
jumps. Self-modifying code can be handled by combining the method with
dynamic translation or emulation. Object code translation can get
within 90% of the speed of native execution on well-behaved programs.
This is the method DEC uses to compile MIPS and Vax code to Alpha.
Torben Mogensen (tor...@diku.dk)
Simon Proven 4th Year CES
>I propose a little experiment. I'll lock you in a room with a PC
>running MS Windows and an Archimedes, but I'll take away the mice from
>both machines. I'll come back after a week, and see which one you're
>using.
Within a day I'd have written a mouse emulator for the Arc's keyboard,
and have that up & running. The PC would then be used as a battering
ram to escape :-)
--
"If you want something done properly, kill Baldrick before you start"
-- Edmund Blackadder
--Simon Proven spr...@cs.strath.ac.uk
William Fang
: Well lets face some facts;
: OS2 requires a minimum of 8Meg Ram to work and a lot more to work
: properly!
Nearly every clone box comes with at least 4M of ram, so the extra
costs of 4MB of ram is Aus$320.
: OS2 takes approximately 40Meg of your hard drive to work at all!
Hard-drive space is running at least than Aus$1.40, (for 340MB IDE)
so the extra requiremnt is Aus$376, which buys you about 4 games
over here, or less than the price of a single application from a
major vendor (eg, Word for Windows, Excel, Lotus 123, etc).
: Not exactly compact is it?
Cost/Benefit ratio probably comes out in OS/2 / 486's favour.
- Jim
R. Brown
Read my post. How much is an ARM chip these days? Plug one into the Emplant
board, and use THAT to run ARMcode. The Amiga can do any necessary graphics
processing, and we'll have RISCOS on a nice draggable screen, thankyou.
Emulating the Amiga on an Arch is OF COURSE possible, but where would you draw
the line? The Arch in it's present form CAN'T use all of the screen modes
available to the Amiga, and emulating a lot of the other hardware would slow
everything to a crawl. An OS level emulation I could cope with, but even then a
lot of programs are looking for a planar screen, and converting that will slow
things down, especially if you're emulating a 68000 too (68020 probably, since
this is now the base Amiga model). HAM modes might also be somewhat difficult.
And the last time any of the people I know looked, you COULDN'T get the
horizontal position of the display beam on current Archs, which you need to
emulate the Copper at any plausible speed... OF course, if you're just going to
emulate the OS, you could re-code the graphics.library (Nice having a modular OS
isn't it (this IS serious. No flame (to the Arch) intended)), and tell it to
refuse to open HAM screens, make all screens non-draggable, etc. but this would
cause at least a good percentage of programs to refuse to run. Hardware level
emulation would be A LOT more difficult, and would CRAWL. Ok, so plug in an Amiga
chipset to help? Well, you've already covered most of the cost of an Amiga. Put a
68020 on the card too, and effectively it IS an Amiga (020s are REALLY cheap
these days...). Oh yes, doing the disk drives could be fun, too.
What I'm basically trying to say is that whilst it's obviously POSSIBLE to
emulate an Amiga on an Arch, it's probably much easier/faster/cheaper to just buy
an Amiga. However, it WOULD be possible to emulate an Arch on an Amiga.
An OS level emulation in either direction would be easy. Hardware would be much
more difficult. Modular operating systems make that sort of thing quite easy.
(Actually, I've got an Arch owner sat next to me here who thinks he might have a
go at Amiga emulation if I give him a hand. I may have to apologise for all of
the above... :-)). Actually... anyone want to try the Arch on an Amiga? (Mind you,
we're only talking OCS/1.3 emulation for the Arch, which is hardly sort of state
of the art. A bit like emulating an Arch with an ARM2 and Arthur. Actually, why
is it both Amiga and Arch owners don't like ANY desktop that's dark blue? :-) :-))
This whole argument could change with the next release of the Amiga operating
system (except perhaps the disks *grin*), which is apparently re-targettable in
the graphics department.
The thing about the 486 emulation came from an announcement from Utilities
Unlimited, who're WRITING a PC emulator for their card. Apparently it was going
to have a 486 in it, but they decided they didn't need it! (*I* thought it was
unbelievable too, but it's not ME that's saying it...)
Don't get me wrong. I *LIKE* the Arch. If the Amiga didn't exist, I'd probably
have one. The two machines are probably the most 'original' platforms currently
on the market! So why don't we all stop shouting at each other? :-)
Robert Brown.
C.P. Brown
|> >Just thought I'd make the pint: The Amiga's the emulation specialist. Give us an
|> >ARM CPU for Emplant (Or let the guys at UU have a go at emulating it. They seem
|> >to think they can get an almost full speed 486 emulation out of a 68040...<grin>)
|> >and a set of RISCOS ROMS, and we'll probably manage. Now, what would the Arch
|> >need for an Amiga emulation?
|>
|> It's more plausible that an Amiga could be emulated on an Arc than
|> vice-versa, I feel. Blitting is easier to do in software on an ARM
|> than ARMcode is to do in a blitter. (-8
I have to disagree here. Whilst it would probably be possible to emulate an Amiga
on an Arch (if you can manage to read the disks :-) ), it would run intolerably
slowly. There is much more to the Amiga's graphics hardware than the blitter.
There is the copper, which provides multiple resoloutions and palettes, and HAM
mode, which is drmatically different to standard video modes. You also have to
bear in mind that the Amiga's screen display is planar, and individual bit planes
can be moved around independantly of the others, providing interesting effects
such as dual playfield. To faithfully emulate the AGA chipset (that is repoduce
the same screen image, in the correct colours and correct resoloution) would
require a display capable of 1400*600 and 800*1100, both in 24 bit. This would
run intolerably slowly on current microprocessors. Emulating an A500 on a fast
Arc could be plausible, but not nice to program. On the other hand, and Amiga
emulating an Arch mainly has the problem of emulating the ARM processor, and I
believe that a 68040 could probably emulate ARM code at a fairly reasonable speed
(comparable to an ARM 2 perhaps?)
|>
|> I don't believe that a 486 could be emulated at more than about a
|> tenth of full speed on a 68040.
|>
|> --Clive.
--
M.D.F. Box
>
>I propose a little experiment. I'll lock you in a room with a PC
>running MS Windows and an Archimedes, but I'll take away the mice from
>both machines. I'll come back after a week, and see which one you're
>using.
>
>Ol
>--
>Trained in Cambridge as an urban attack cyclist.
>
That is blatantly stupid. It's like saying I'll take away the keyboard and see which one you use. The mouse is an integral part of RiscOS, and all systems come with one. The only reason MSh*t felt it necessary to put in a (lousy) keyboard control system is that the PC doesn't come with a mouse, it is an optional extra (albeit a normally supplied one). And when Windoze was originally designed (don't make me laugh) mice were few and far between.
Malcolm Box
M.D....@newcastle.ac.uk
No .sig : Save electrons today for a better tommorrow
| Hartley's Law:
| You can lead a horse to water,
M.D....@newcastle.ac.uk | but if you can get it to float on its
| back, then you've got something.
Oliver Betts
>>I propose a little experiment. I'll lock you in a room with a PC
>>running MS Windows and an Archimedes, but I'll take away the mice from
>>both machines. I'll come back after a week, and see which one you're
>>using.
In article <CHD4C...@newcastle.ac.uk>,
M.D.F. Box <M.D....@newcastle.ac.uk> wrote:
>That is blatantly stupid. It's like saying I'll take away the keyboard
and see which one you use. The mouse is an integral part of RiscOS, and
all systems come with one. The only reason MSh*t felt it necessary to
put in a (lousy) keyboard control system is that the PC doesn't come
with a mouse, it is an optional extra (albeit a normally supplied one).
And when Windoze was originally designed (don't make me laugh) mice were
few and far between.
First of all, wrap your lines if you think they are worth reading.
Ideally at something a little less than 80 columns, so they can get
indented and still be readable by most people.
Secondly, I realise that this is c.s.a.*advocacy* and that therefore
facts aren't allowed usually allowed to intrude, but just for once:
Choosing everything from menus slows you up. Few experienced users
use Selection.Copy, they use Ctrl-C. If you're typing in text, it's
annoying to have to move your hand to the mouse.
I'm not in love with MS Windows (I much prefer my ARM3 Arc to this 66MHz
486DX2 I'm typing on now), but it does feel much more coherently
designed than RISC OS. The most obvious example is that you can do
almost everything without removing your hands from the keyboard.
Various filer hacks and patches are around which add key shortcuts to
the RISC OS filer, so users seem to want this.
Incidentally, you say:
"The mouse is an integral part of RiscOS, and all systems come with one"
What about the A4?
Admittedly it has a keyboard mouse simulator, which is good -
considering it is a keyboard mouse simulator.
Olly
A.W. Oakden
> costs of 4MB of ram is Aus$320.
>
> so the extra requiremnt is Aus$376, which buys you about 4 games
> over here, or less than the price of a single application from a
> major vendor (eg, Word for Windows, Excel, Lotus 123, etc).
>
>
>
OS2 still takes more Ram than the machine comes with as standard!
Also OS2 is unlikely to take hold of the market given the head start
handed to MicroSoft by IBM. PC compatables will probably be lumbered with
Windows for many years to come.
As far as Cost/Benefit ratio goes - exactly what benefit are you
refering to? If you mean OS2 Vs Windows ;
I do not know what the
compatability situation is between them, but, given the problems with
getting software to work on windows I should guess that much rewriting
of software will be needed to support OS2. This should cancel out any
real advantage that the new product has (at least in the eyes of upper
managament who buy the machines).
If the Cost/Benefit ratio to which you refer is between the PC and the
Arc then there really is no comparason, The two machines are entirely
different both in architecture and the uses to which they are put. I
would rather pay more for the elegant design of the Arc that put up
with the 486.
Tony
Kyle Dawkins
>> Nearly every clone box comes with at least 4M of ram, so the extra
>> costs of 4MB of ram is Aus$320.
>>
>> Hard-drive space is running at least than Aus$1.40, (for 340MB IDE)
>>
>>
>>
>I think this is irrelevent!
>
>OS2 still takes more Ram than the machine comes with as standard!
8MB clones are extremely common and IBM bundles all its machines
with this much RAM and OS/2 here now.
>Also OS2 is unlikely to take hold of the market given the head start
>handed to MicroSoft by IBM. PC compatables will probably be lumbered with
>Windows for many years to come.
>
My money is on OS/2. Here, you can buy OS/2 for Windows
for $70 (cheaper if you have a CD ROM drive). That's
35 quid for you pommies. Now, I think your average game is
that much. OS/2 installs itself on top of Windows
and DOS and does NOT take up 40 MB... up to about
35 if you choose to install Windows and DOS with it.
Nobody does... they already have Windows and DOS.
Besides, I am not saying that OS/2 is the be-all and end-all
of OS's either, just that I am fed up with
people claiming that PCs and Windows are one and the same.
>
For those of you who don't know, OS/2 will run Windows software
and DOS software concurrently with OS/2 software... not that
appeals to me greatly since I think there is some wonderful
software written for each. Then again, it may not appeal
to you. So you can always choose another OS.
>Tony
Cheers
Kyle
PS. Our mailer is screwed... my address is daw...@music.mcgill.ca
Russell Skingsley
>I propose a little experiment. I'll lock you in a room with a PC
>running MS Windows and an Archimedes, but I'll take away the mice from
>both machines. I'll come back after a week, and see which one you're
>using.
What kind of an argument is this?
What would this prove?
The mouse is an integral part of RISC OS operation so what?
Why not say you'll put a PC running DOS and a MAC running system 7 in a room
and take away the mouse. Obviously after a week they'll be using the DOS
machine so it must have a better interface than the MAC!
>Ol
>--
>Trained in Cambridge as an urban attack cyclist.
##############################################################################
Russell Skingsley Email:
Guided Weapons Division r...@homer.dsto.gov.au
Defence Science and Technology Organisation
"Since I gave up hope I feel a lot better" - Steve Taylor
##############################################################################
Clive Jones
>Read my post. How much is an ARM chip these days? Plug one into the Emplant
>board, and use THAT to run ARMcode. The Amiga can do any necessary graphics
>processing, and we'll have RISCOS on a nice draggable screen, thankyou.
Depends which one you want, really.
You don't appear to be talking about emulation after all. It would
seem that this Emplant board (of which I know nothing, not being an
Amiga owner!) takes foreign CPU chips and uses them to provide
cross-compatability. This is not remotely the same as an emulator, and
if you want to see what's possible along these lines with an Arc you
should be looking at Aleph-1's 486 PC card. This, incidentally, does
emulate the Windows PLI on the ARM to accelerate things a little.
>Emulating the Amiga on an Arch is OF COURSE possible, but where would you draw
>the line? The Arch in it's present form CAN'T use all of the screen modes
>available to the Amiga, and emulating a lot of the other hardware would slow
>everything to a crawl. [...]
I never said it was possible to emulate an Amiga at any reasonable
speed on an Arc, I merely said it would be more viable than an Arc
emulator on an Amiga.
> (Nice having a modular OS
>isn't it (this IS serious. No flame (to the Arch) intended))
I see. You think the Arc doesn't have a modular OS? What's wrong with:
SWI OS_Claim, WrchV, address of routine, private data handle
...to replace the graphics output routines, pray tell?
> However, it WOULD be possible to emulate an Arch on an Amiga.
I don't believe I have deleted any comments of yours that justify this
statement - it would be just as exciting watching an Amiga trying to
emulate the Arc's sound capabilities as it would be to watch an Arc
trying to emulate Amiga HAM modes, for example.
--Clive.
Clive Jones
> On the other hand, and Amiga
>emulating an Arch mainly has the problem of emulating the ARM processor, and I
>believe that a 68040 could probably emulate ARM code at a fairly reasonable speed
>(comparable to an ARM 2 perhaps?)
Well, given that Sun Sparcstations can't manage 1MIPS, it would be
quite clever to manage 5MIPS on a 68040...
I agree that the ARM is a more significant proportion of the task of
emulating an Arc than the 680x0 is of emulating an Amiga, though.
--Clive.
Michael Williams
>It would be interesting to hear how fast and complete this prototype
>emulation is. Is it only ARM code, or is the full Arc hardware
>simulated?
ARM code is already emulated, at reasonable speed, by ARM's own
ARMulator. This is shipped as part of the toolkit. (It is integrated
into the debugger, allowing you to run the debugger without any
hardware, on a fully emulated ARM.) One problem with this approach is
that it needs to be quite a good emulation, since you need to be able
to single-step it, set breakpoints etc. Still, on a SPARC IPC I get
about a quarter of a MIP out under emulation.
>With good hardware simulation, OS simulation is reduced to emulating
>the instructions and hardware (as the current Acorn PC emulator does).
>This is, however, a slow method. A faster metod is to rewrite (parts
>of) the OS in native code, and catch the calls to the OS. This is the
>approach in WABI and for running Macintosh programs on the PowerPC.
This is also the approach taken in ARMulator.
>Instruction set simulation is the easiest part. Any good programmer
>can over a weekend make a simulation of an instruction set. IT takes a
>bit more effort to make it fast.
>
>A traditional emulation simulates each instruction on its own. A naive
>approach will compare the op-code field of the instruction with the
>various possibilities, and then decode the operand fields and perform
>the relevant action. This is, however, very slow. An average of about
>100 instructions is needed to emulate a single instruction. Various
>methods can improve on this.
>
>If the instruction word is small, e.g. 8 or 16 bit a jump table can be
>used.
Most C authors would use a switch statement to do this, and the
compiler will try its best to convert this into a branch table. There
are various tricks for making ARM emulation go faster, none of which
I'm going to reveal here 8-).
>memory. Let's say that an 680x0 instruction requires an average of 5
>ARM instructions to emulate, plus one word in the jump table. This
>requires 1.5MB of memory, so it is out of the question on small
>machines, but perfectly feasible on 4MB Arcs.
I think 5 ARM instructions per. 680x0 instruction is being a little
optimistic.
____________________________________________________________________________
\ ^ / Michael Williams Advanced RISC Machines Limited
|\/|\/\ mwil...@armltd.co.uk Swaffham Bulbeck, Cambridge, UK
| |(__)Cymdeithas Genedlaethol Traddodiad Troi Teigrod Mwythus Ben I Waered
"I might well think that Matti, ARM Ltd. couldn't possibly comment."
Andrew J D Hurley
>The mouse is an integral part of RISC OS operation so what?
>
>Why not say you'll put a PC running DOS and a MAC running system 7 in a room
>and take away the mouse. Obviously after a week they'll be using the DOS
>machine so it must have a better interface than the MAC!
This is a stupid thread but it's quite a laugh.
How about a MAC and an Arc (both without mice), I suspect the Arc would
win here since it would be easy to knock up mouse emulator and anyway
you don't NEED a mouse because there's a CLI built in.
-----------------------------------------------------------------------
Andrew J D Hurley , <>< | Email: A.J.D....@bnr.co.uk
| or Andrew.Hu...@bnr.ca
BNR Europe Ltd., London Road, | route: uunet!ukc!bnr!ajdh
Harlow, Essex CM17 9NA, UK. | Phone: +44 279 402535
Robin Watts
Olly wrote:
>>I propose a little experiment. I'll lock you in a room with a PC
>>running MS Windows and an Archimedes, but I'll take away the mice from
>>both machines. I'll come back after a week, and see which one you're
>>using.
Taking the bait M.D.F. Box replied:
>That is blatantly stupid. It's like saying I'll take away the keyboard
>and see which one you use. The mouse is an integral part of RiscOS, and
>all systems come with one. The only reason MSh*t felt it necessary to
>put in a (lousy) keyboard control system is that the PC doesn't come
>with a mouse, it is an optional extra (albeit a normally supplied one).
>And when Windoze was originally designed (don't make me laugh) mice were
>few and far between.
First of all, wrap your lines if you think they are worth reading.
Ideally at something a little less than 80 columns, so they can get
indented and still be readable by most people.
Indeed. I wish more people would do this. About 60 columns
is reasonable for posts in likely-to-be-cascaded discussions
like this. (Though Ikki takes it a bit far :-) )
Secondly, I realise that this is c.s.a.*advocacy* and that therefore
facts aren't allowed usually allowed to intrude, but just for once:
Choosing everything from menus slows you up. Few experienced users
use Selection.Copy, they use Ctrl-C. If you're typing in text, it's
annoying to have to move your hand to the mouse.
True. However, keyboard shortcuts should not be implemented to
the exclusion of all else. To Paraphrase Tog (Tog on Interface)
'Here at Apple we have done a cool 50 million dollars worth
of R & D . The results: Users claim that Keyboard shortcuts
improve productivity, where research consistently shows that
they actually lower productivity.'
Keyboard shortcuts are great if they are standard, and only a
small set (Like C-z, C-x, C-c, C-v, f3, f4) but obscure ones
only slow you down. To insist (as MS do) that *all* apps
must be drivable purely from the keyboard is insane. This
inevitably detracts from the usability of a system.
I'm not in love with MS Windows (I much prefer my ARM3 Arc to this 66MHz
486DX2 I'm typing on now), but it does feel much more coherently
designed than RISC OS.
Ooohh.... I can feel an expletive welling up within me. MS Windows
(in common with most PC stuff) was not designed. It has 'evolved'
from a distinctly dodgy start towards a state where everything you'd
expect to be there is there, but in odd places/ways due to the
enforced backwards compatibility of everything.
The most obvious example is that you can do
almost everything without removing your hands from the keyboard.
This dates from the earliest versions which were designed for
use without a mouse.
Various filer hacks and patches are around which add key shortcuts to
the RISC OS filer, so users seem to want this.
Indeed, but of these shortcuts only C-x, C-r, and f3 are ever used
by users (Delete, Rename, Copy). Anything else it is faster to do by
the mouse.
Incidentally, you say:
"The mouse is an integral part of RiscOS, and all systems come with one"
What about the A4?
what about it? You get a mouse with that. An alternative system is
included, but a mouse is supplied.
Olly
--
Trained in Cambridge as an urban attack cyclist.
Robin
C.P. Brown
|> I don't believe I have deleted any comments of yours that justify this
|> statement - it would be just as exciting watching an Amiga trying to
|> emulate the Arc's sound capabilities as it would be to watch an Arc
|> trying to emulate Amiga HAM modes, for example.
Not sure about this. AFAIK, the major difference between thw two is the fact that
the Arch has 8 channels, wheras the Amiga has 4. However, emulating 8 channels on
a 4 chennel sound chip is not a processor intensive task. Take a look at the
Amiga's OctaMED program, which on my A1200 provides 8 chennel sound by
sacrificing a negligible amount of CPU time. As a side note, apparently the
forthcomming AAA chips sound capibilities are 8 channels of 16 bit sound at
100KHz. Should sound very nice indeed... (I want one :-) )
C.P. Brown
better interface than RiscOS is frankly rather silly. Yes, Windows has lots of
keyboard shortcuts, but I find that most users tend to have to resort to them
becuase Windows mouse interface is akward and clumsy. Just some of the things I
really hate about Windows mouse interface:
- The menus stay there if you let go of the button (for someone used to using
the Amiga's menuing system, this is particually irritating)
- A window cannot be moved to the back.
- The active window is, by definition, the frontmost one.
- Floppy disk access seems to be more important than the ability to move the
mouse pointer.
These are the 'features' of MS-Windows that I have found, in use, to make the
interface restrictive and akward. Don't get me wrong, I like hot key shortcuts,
and have several set up on my own system, but there are lots of things in a well
designed interface that a mouse is more suited to, and IMHO, Windows is very weak
in this respect.
A G Jackson
|> In article <2dj6q9$h...@taki.nsict.org>, cl...@nsict.org (Clive Jones) writes:
|>
|> |> I don't believe I have deleted any comments of yours that justify this
|> |> statement - it would be just as exciting watching an Amiga trying to
|> |> emulate the Arc's sound capabilities as it would be to watch an Arc
|> |> trying to emulate Amiga HAM modes, for example.
|>
|> Not sure about this. AFAIK, the major difference between thw two is the fact that
|> the Arch has 8 channels, wheras the Amiga has 4. However, emulating 8 channels on
|> a 4 chennel sound chip is not a processor intensive task. Take a look at the
|> Amiga's OctaMED program, which on my A1200 provides 8 chennel sound by
|> sacrificing a negligible amount of CPU time. As a side note, apparently the
|> forthcomming AAA chips sound capibilities are 8 channels of 16 bit sound at
|> 100KHz. Should sound very nice indeed... (I want one :-) )
Of course, there is a soundtracker program for the Arc which provides 16 channel
sound by sacrificing a negligible amount of CPU time...
Adrian
+--------------------------------------------------------------------------+
| Adrian Jackson | a.g.j...@durham.ac.uk | 27 North Bailey, DURHAM |
+--------------------------------------------------------------------------+
| Holly: Ahead groove factor 5! Yeah! |
+--------------------------------------------------------------------------+
R. Brown
|> In article <1993Dec1.1...@infodev.cam.ac.uk> Robert Brown <rb1...@hermes.cam.ac.uk> writes:
|> >Read my post. How much is an ARM chip these days? Plug one into the Emplant
|> >board, and use THAT to run ARMcode. The Amiga can do any necessary graphics
|> >processing, and we'll have RISCOS on a nice draggable screen, thankyou.
|>
|> Depends which one you want, really.
|>
|> You don't appear to be talking about emulation after all. It would
|> seem that this Emplant board (of which I know nothing, not being an
|> Amiga owner!) takes foreign CPU chips and uses them to provide
|> cross-compatability. This is not remotely the same as an emulator, and
|> if you want to see what's possible along these lines with an Arc you
|> should be looking at Aleph-1's 486 PC card. This, incidentally, does
|> emulate the Windows PLI on the ARM to accelerate things a little.
The Emplant card is basically designed as a multi-platform emulation system for
the Amiga. It CAN take multiple processors and do things that way, but the
designers seem to be aiming more at software emulation of the processor, so that
it is possible to multitask it. I wouldn't be quite so quick to dismiss their
cliams (re the 486 emulation mentioned earlier), since they've currently produced
a Mac emulation that multi-tasks with AmigaDOS using the same processor (i.e.
whatever's built into the machine), a task which was hitherto thought to be
impossible (due to lots of interesting and strange things, not least of which is
that NEITHER operating system was designed to do this...)
|> >Emulating the Amiga on an Arch is OF COURSE possible, but where would you draw
|> >the line? The Arch in it's present form CAN'T use all of the screen modes
|> >available to the Amiga, and emulating a lot of the other hardware would slow
|> >everything to a crawl. [...]
|>
|> I never said it was possible to emulate an Amiga at any reasonable
|> speed on an Arc, I merely said it would be more viable than an Arc
|> emulator on an Amiga.
In some ways, I'll still dispute this. The only 'nasty' part of getting an Arch
emulation working on the Amiga is the processor, as you say below. However, to
get the Arch to emulate an Amiga you would have to either emulate most of the
low-level hardware (and as far as I'm aware there's a LOT more of that in the
Amiga than in the Arch), or go for an OS-only emulation.
I get the feeling that whichever way you try the emulation, without a significant
amount of assisting hardware, a reasonably compatible emulation is going to be
VERY slow compared to the original.
|> > (Nice having a modular OS
|> >isn't it (this IS serious. No flame (to the Arch) intended))
|>
|> I see. You think the Arc doesn't have a modular OS? What's wrong with:
|> SWI OS_Claim, WrchV, address of routine, private data handle
|> ...to replace the graphics output routines, pray tell?
NO NO NO NO NO!
I WASN'T being at all nasty to the Arch! Maybe I should have stated things more
clearly (!!). I REALISE THE ARCH HAS A MODULAR OPERATING SYSTEM! So does the
Amiga. That should make anybody's attempts at OS emulation MUCH easier. I made
that comment about certain OTHER operating systems that haven't worked out how to
be modular yet...
|> > However, it WOULD be possible to emulate an Arch on an Amiga.
|>
|> I don't believe I have deleted any comments of yours that justify this
|> statement - it would be just as exciting watching an Amiga trying to
|> emulate the Arc's sound capabilities as it would be to watch an Arc
|> trying to emulate Amiga HAM modes, for example.
|>
|> --Clive.
Hmmmm. Somehow I think I could do without the sound more easily than I could do
without the display... :-)
^^^ - Yes, this was a joke. PLEASE don't take it seriously.
Seriously, though, I seem to remember the Arch has 8-channel sound? (16/8 bit
doesn't REALLY matter. You'll take a nasty drop in quality, but that's not that
much of a problem. Have you heard the A4? :-)) Well, I've had 8-channel sound
packages on the Amiga for ages.
Anyway, as I said before: don't take any of this too seriously. I'm quite willing
to be corrected on any of the points I've made. And I'd LOVE to see BOTH an Amiga
emulator for the Arch (That'd be one up for you on the PC crowd), AND an Arch
emulator for the Amiga. I just don't think we're likely to see it in the
foreseeable future...
Robert Brown.
Oliver Betts
Robin Watts <ct9...@black.ox.ac.uk> wrote:
>In article <2digr1$n...@news.mantis.co.uk> ol...@mantis.co.uk (Oliver Betts) writes:
>
> Olly wrote:
> >>I propose a little experiment. I'll lock you in a room with a PC
> >>running MS Windows and an Archimedes, but I'll take away the mice from
> >>both machines. I'll come back after a week, and see which one you're
> >>using.
>
> Taking the bait M.D.F. Box replied:
Are you implying that I'm stirring things up? Hmmm? Okay, maybe a little :)
>Indeed. I wish more people would do this. About 60 columns
>is reasonable for posts in likely-to-be-cascaded discussions
>like this. (Though Ikki takes it a bit far :-) )
Methinks he has a 40 column terminal. Incidentally, has anyone else
wondered why the number of '=' in Amit Gupta .sig varies from day to
day?
>> [..]
>
>True. However, keyboard shortcuts should not be implemented to
>the exclusion of all else. To Paraphrase Tog (Tog on Interface)
>'Here at Apple we have done a cool 50 million dollars worth
>of R & D . The results: Users claim that Keyboard shortcuts
>improve productivity, where research consistently shows that
>they actually lower productivity.'
Ah, but that's Mac users. There's a quote from in !Yow from John
Cook at Apple: (from memory)
"Our research shows that many users find double-clicking
a hard concept to grasp."
>Keyboard shortcuts are great if they are standard, and only a
>small set (Like C-z, C-x, C-c, C-v, f3, f4) but obscure ones
>only slow you down. To insist (as MS do) that *all* apps
>must be drivable purely from the keyboard is insane. This
>inevitably detracts from the usability of a system.
If you can't remember the shortcut, then you can press Alt and use the
cursor keys to operate the menus. This is the major problem with this
idea on RISC OS - this is fine for pull-down menus, but with pop-up ones
it doesn't work - which menu do you bring up?
BTW by pop-up, I don't mean menus on a button - I mean menus brought
up by pressing the Menu button on a window. There's probably a better
word, but it escapes me right now.
Research seems to show that the problem with pop-up menus is that they
are less obvious. Indeed with a few freeware apps I've only discovered
some menus after much use. The major advantage with them is that you
don't need to move the mouse much to get menus up.
>Ooohh.... I can feel an expletive welling up within me. MS Windows
>(in common with most PC stuff) was not designed. It has 'evolved'
>from a distinctly dodgy start towards a state where everything you'd
>expect to be there is there, but in odd places/ways due to the
>enforced backwards compatibility of everything.
Okay, put your hand on your heart and say:
"The Arthur desktop was well designed and usable."
[And try not to giggle].
MS Windows is built on DOS, which is a programmers nightmare. Great - I
have 32 Mb in my machine, but my C program can use 600k max, and to get
anywhere near that you spend days fiddling with what gets loaded high.
To get more I have to write XMS/EMS memory stuff, or compile with a DOS
extender, which won't run under Windows. The major problem programming
Windows itself (like I'm meant to be at the moment) is the size of the
manuals. There are functions to do things no sane programmer would
want, so you have to wade a lot to find what you want.
>[...]
>Indeed, but of these shortcuts only C-x, C-r, and f3 are ever used
>by users (Delete, Rename, Copy). Anything else it is faster to do by
>the mouse.
Depends how much you use the software - I can perform any common (to me)
operation in !Zap using the keyboard, for eg. The only thing I can
think of I use the menus for is where you need to (eg changing the
indent or width). That's at least 40 shortcuts, I reckon.
Ol
Oliver Betts
C.P. Brown <cpb...@hermes.cam.ac.uk> wrote:
>All this talk about the ability to use Windows without a mouse making somehow a
>better interface than RiscOS is frankly rather silly. Yes, Windows has lots of
>keyboard shortcuts, but I find that most users tend to have to resort to them
>becuase Windows mouse interface is akward and clumsy. Just some of the things I
>really hate about Windows mouse interface:
>
> - The menus stay there if you let go of the button (for someone used to using
>the Amiga's menuing system, this is particually irritating)
Simply incorrect.
Okay. Press left button on menu, and hold it down. Go down to option,
let go and option is chosen.
Alternatively, click left button on menu. It pops up and clicking again
chooses an option.
The other two possible behaviours (drag and menu stays up when you
release) and (menu disappears if you just click on it) aren't terribly
useful. The Arc does the second, and it forces you to drag - boy is it
annoying when you're trying to operate the mouse on something too small
- eg a beer mat - you keep letting go by mistake and choosing things.
You gain nothing by loosing either behaviour.
>[..]
>
> - Floppy disk access seems to be more important than the ability to move the
>mouse pointer.
The Arc has a hardware driven pointer, and so cheats. Floppy accessing makes
most computers chuggy until it's finished. Giving the pointer priority would
make it take longer.
>These are the 'features' of MS-Windows that I have found, in use, to make the
>interface restrictive and akward. Don't get me wrong, I like hot key shortcuts,
>and have several set up on my own system, but there are lots of things in a
>well
>designed interface that a mouse is more suited to, and IMHO, Windows is very
>weak
>in this respect.
Blimey, I'm championing Windows - must be at a loose end today...
Oliver Betts
Andrew J D Hurley <aj...@bnr.co.uk> wrote:
>In article <rws.42....@homer.dsto.gov.au> r...@homer.dsto.gov.au (Russell Skingsley) writes:
>>Why not say you'll put a PC running DOS and a MAC running system 7 in a room
>>and take away the mouse. Obviously after a week they'll be using the DOS
>>machine so it must have a better interface than the MAC!
>
>This is a stupid thread but it's quite a laugh.
Why thankyou... :)
>How about a MAC and an Arc (both without mice), I suspect the Arc would
>win here since it would be easy to knock up mouse emulator and anyway
>you don't NEED a mouse because there's a CLI built in.
Looks like the moral of this thread is:
"Don't get locked in a room with a Mac"
C.P. Brown
|> In article <1993Dec2.1...@infodev.cam.ac.uk>,
|> C.P. Brown <cpb...@hermes.cam.ac.uk> wrote:
|> >All this talk about the ability to use Windows without a mouse making somehow a
|> >better interface than RiscOS is frankly rather silly. Yes, Windows has lots of
|> >keyboard shortcuts, but I find that most users tend to have to resort to them
|> >becuase Windows mouse interface is akward and clumsy. Just some of the things I
|> >really hate about Windows mouse interface:
|> >
|> > - The menus stay there if you let go of the button (for someone used to using
|> >the Amiga's menuing system, this is particually irritating)
|>
|> Simply incorrect.
|>
|> Okay. Press left button on menu, and hold it down. Go down to option,
|> let go and option is chosen.
|>
|> Alternatively, click left button on menu. It pops up and clicking again
|> chooses an option.
presumably, when you click once, you let go of the button, right? The menu stays
there... What if you dont want to select anything after all? You spend time
moving the mouse to a neutral area of the screen, before clicking again..
Hardly optimal. What also irritates me is that Windows uses the select button
for menu's. Now (from an Amiga point of view, because that's what I have),
this is very annoying, because it means you can't make multiple selections
from menus.
|>
|> The other two possible behaviours (drag and menu stays up when you
|> release) and (menu disappears if you just click on it) aren't terribly
|> useful. The Arc does the second, and it forces you to drag - boy is it
|> annoying when you're trying to operate the mouse on something too small
|> - eg a beer mat - you keep letting go by mistake and choosing things.
Well, I never experience this, because I have an optical mouse, which doesn't
work terribly well on a beer-mat :-) I also don't tend to accidentally select
options.. maybe because my mouse buttons are good quality...
|>
|> You gain nothing by loosing either behaviour.
|>
|> >[..]
|> >
|> > - Floppy disk access seems to be more important than the ability to move the
|> >mouse pointer.
|>
|> The Arc has a hardware driven pointer, and so cheats. Floppy accessing makes
|> most computers chuggy until it's finished. Giving the pointer priority would
|> make it take longer.
Why should floppy disk access make the machine chuggy. When I access the floppy
on my machine (again, this is from an Amiga perspective), it certainly doesn't
seem chuggy, and all my programs continue to run with no noticible slowdown.
(This problem on a PC basically comes down to a poorly designed architecture)
|>
|> >These are the 'features' of MS-Windows that I have found, in use, to make the
|> >interface restrictive and akward. Don't get me wrong, I like hot key shortcuts,
|> >and have several set up on my own system, but there are lots of things in a
|> >well
|> >designed interface that a mouse is more suited to, and IMHO, Windows is very
|> >weak
|> >in this respect.
|>
And do you have any comments about my other points? (no way to send a window to
the back, and the active window must be at the front)
Hugo Fiennes
>Not sure about this. AFAIK, the major difference between thw two is the fact
> that
>the Arch has 8 channels, wheras the Amiga has 4. However, emulating 8 channels
> on
>a 4 chennel sound chip is not a processor intensive task. Take a look at the
>Amiga's OctaMED program, which on my A1200 provides 8 chennel sound by
>sacrificing a negligible amount of CPU time. As a side note, apparently the
>forthcomming AAA chips sound capibilities are 8 channels of 16 bit sound at
>100KHz. Should sound very nice indeed... (I want one :-) )
Emulating 8 channels on a 4 channel sound chip takes all the amiga's cpu
time - witness octamed on an A500 (total freeze during playing, not even
the mousepointer moves - OK, so maybe this is partly by design, so you can't
interfere, but it isn't doing much useful apart from playing music).
I believe the A1200 has hardware support for 8 channels, which would explain
it.
(Octamed on an a500 sounds awful because you get max 14kHz sample rate per
channel - yuck!).
Hugo
--
Hugo "I've suffered a fatal internal error and must exit immediately" Fiennes
alt...@cryton.demon.co.uk / vox +44 749 670058 / bbs +44 749 670030 2:252/102
----------------------------- "...and that was my last treated digestive too"
C.P. Brown
|> In article <1993Dec2.1...@infodev.cam.ac.uk> cpb...@hermes.cam.ac.uk writes:
|>
|> >Not sure about this. AFAIK, the major difference between thw two is the fact
|> > that
|> >the Arch has 8 channels, wheras the Amiga has 4. However, emulating 8 channels
|> > on
|> >a 4 chennel sound chip is not a processor intensive task. Take a look at the
|> >Amiga's OctaMED program, which on my A1200 provides 8 chennel sound by
|> >sacrificing a negligible amount of CPU time. As a side note, apparently the
|> >forthcomming AAA chips sound capibilities are 8 channels of 16 bit sound at
|> >100KHz. Should sound very nice indeed... (I want one :-) )
|>
|> Emulating 8 channels on a 4 channel sound chip takes all the amiga's cpu
|> time - witness octamed on an A500 (total freeze during playing, not even
|> the mousepointer moves - OK, so maybe this is partly by design, so you can't
|> interfere, but it isn't doing much useful apart from playing music).
OctaMED does not take all CPU time on an A500. It takes a noticable chunk of it,
but the mouse pointer moves quite freely. The probable reason it locked when you
tried it is because OctaMED uses interrupts to play it's sound, and if you run it
in 8 channel mode, on a machine with only CHIP RAM, in a high bus-contention
graphics mode, then the machine spends all it's time servicing interrupts, and
cannot run any other tasks. Your comment about not being able to interfere is
silly. OctaMED allows you to edit scores whilst they are playing, as well as
use most of the other features of the program. Try it on a mcahine with FASTRAM.
You should see a noticable speedup.
|>
|> I believe the A1200 has hardware support for 8 channels, which would explain
|> it.
You believe wrong. The A1200 has the same 4 channel sound as the A500. The reason
it runs faster is because the A1200 is a faster machine!
|> (Octamed on an a500 sounds awful because you get max 14kHz sample rate per
|> channel - yuck!).
You should see it on an A1200. it sounds perfect, and takes up very little extra
CPU time over playing 4 channels (somewhere in the region of 2%)
R. Brown
|>
|> >Not sure about this. AFAIK, the major difference between thw two is the fact
|> > that
|> >the Arch has 8 channels, wheras the Amiga has 4. However, emulating 8 channels
|> > on
|> >a 4 chennel sound chip is not a processor intensive task. Take a look at the
|> >Amiga's OctaMED program, which on my A1200 provides 8 chennel sound by
|> >sacrificing a negligible amount of CPU time. As a side note, apparently the
|> >forthcomming AAA chips sound capibilities are 8 channels of 16 bit sound at
|> >100KHz. Should sound very nice indeed... (I want one :-) )
|>
|> Emulating 8 channels on a 4 channel sound chip takes all the amiga's cpu
|> time - witness octamed on an A500 (total freeze during playing, not even
|> the mousepointer moves - OK, so maybe this is partly by design, so you can't
|> interfere, but it isn't doing much useful apart from playing music).
Ahem. It DOESN'T freeze up completely, thank you very much (unless, perhaps, this
computer that I've been using all this time with 'A500' written on it is actually
NOT an A500 after all...). If you want to increase the responsiveness, set the
number of colours used by OctaMED lower. By default it uses 50% of the processor
bandwidth for it's display! (I KNOW this is ridiculous, but the machine was,
basically, designed in 1984, so it's not THAT bad...). On a 1200 there is NO
noticeable slowdown.
|> I believe the A1200 has hardware support for 8 channels, which would explain
|> it.
Interesting. Since they use EXACTLY THE SAME CHIP, does the A500 have 8 channels
too? Serously, though, the A1200 has 4 channels, same as the 500. It just has
more processor time, and can therefore do 'HQ' mode with 8 channels.
|> (Octamed on an a500 sounds awful because you get max 14kHz sample rate per
|> channel - yuck!).
Use less sound channels and switch HQ on. And I don't think it's that bad,
actually, but I've no figures to quote you. Anyway, as I said above, it's not bad
for a vintage 1984 machine, and they don't even MAKE 500s anymore (haven't for
over a year now). The A1200 is the base Amiga machine now (Since production of
the A600 is, apparently, being wound down too).
|> Hugo
|>
|> --
|> Hugo "I've suffered a fatal internal error and must exit immediately" Fiennes
|> alt...@cryton.demon.co.uk / vox +44 749 670058 / bbs +44 749 670030 2:252/102
|> ----------------------------- "...and that was my last treated digestive too"
Look, I just wish people would make sure they knew what they were talking about
before they critise someone else's computer. I try to get most of my posts here
about the Arc read over by at least one Arc owner before I send them off, to make
sure I haven't screwed up on what the Arc can/can't do. If other people would
perhaps make an effort to be so careful, maybe less flame wars would develop...
Robert Brown.
Jordan K. Hubbard
So what? These arguments that say "if it can't run in 2MB of memory
and 10MB of hard disk space it's rubbish" are specious at best.
If you want cheap, go buy a Commodore 64. I'm willing to buy the
extra memory and disk space as required, just so long as the software
is available to take advantage of it. If the software ain't there, it
doesn't matter how much I stick in the machine, now does it? I'll take
expensive options over no options at all.
getting software to work on windows I should guess that much rewriting
of software will be needed to support OS2. This should cancel out any
Not really. The Windows emulation is pretty good. If you want to write
special software to take advantage of threads, well then more power to
ya. There are some 3 million OS/2 users now, so it's not like you exactly
have a restricted market to sell into!
Jordan
--
(Jordan K. Hubbard) j...@violet.berkeley.edu, j...@al.org, j...@whisker.lotus.ie
Jordan K. Hubbard
>|> > Um.. [clears throat] I'd kinda like an FPU, myself.. As soon as I can
>|> > find somebody who'll sell me one for my A540, I'm buying one. Any
>|> > suggestions? Everybody seems keener to offer upgrades for the A5000's
>|> > these days. So much for being at the forefront. Sob..
Forefront? What forefront? Acorn have dropped the A540, and in fact
did so ages agoo. Sorry.
How quickly they forget. Just a year ago this *WAS* the "forefront".
That's certainly what they called it when I bought mine, anyway..
And yes, I know it was dropped (though your use of "ages ago" for
a 7 month elapsed period indicates that you must be very young indeed! :-).
D.J.Davies
> DJDa...@cen.ex.ac.uk (D.J.Davies) writes:
>
> > There is not much wrong with the machine only three things stop
> > me from bying one instead of the wonderfull Amiga I have at the
> > moment.......
> >
> > 1) The Machine is too unstable (probably due to badly written
> > software (see point two) ) I've been working with RiscOs 3.1
> > A3000's for a year and a half now.
>
> It is? News to me. My machine does lock up now and then but mostly when I
> have been working on some software or testing some badly written freeware
> software. (And you kinda have to expect trouble with freeware, it one of the
> reasons it is so cheap.)
>
> > Acually I was refering to educational programs (sad old BBC
> > ports in the main), things like impression & Eurika seem to
> > be fairly stable (but only just)
>
> But all the commercial software I use works like a charm.
> > I was trying to use a copy of a piece of software the other
> > day (First Sence) It has a licience of only one copy, so I
> > don't want to install it on a HD every time I use it (on a
> > machine with no HD it works) and (on a machine with an IDEFS
> > drive it does'nt find a directory that is there!)??? On any
> > other computer it would be as simple as putting the disc in the
> > machine HD or not and turning it on, (can you force the Arc to
> > boot from the floppy if there is one, or if not boot from HD)?
>
> > 2) The badly written comercial software, The envoirnment is
> > nice the chip is very fast (why no Co-Processors), of all the
> > packages that I've used only three impressed me.
>
> Which three were these and why do they stand out? Also what is wrong with
Because the three I liked actually worked for a change.....
without drive problems etc...
> the other software? It is all very well to make these claims but a little
> actuall justification for their validity would be nice.
>
> > 3) The filosiphy (I can't spell to save my life) of the Arc is
> > that software is modular If your wordprocessor has'ent got a
> > table editor? then buy one. (What a pain in the Ass!)
>
> And you would prefer monolithic lumbering packages that chew HD space and do
> everything? Well if that works for you no problems from me, but I kinda prefer
> the modularity myself. For reasons I can go into if you like but they should be
> obvious if you have done computer science.
They are only lumbering on the PC I have a WP for the Amiga
that is exactly like Ami Pro (with every function you can think
of) (Final writer) and yet the prog only takes 500k and whole
package including 120 postscript type 1 fonts and 100 colour
eps images take 5 megs
>
> > I would buy an archimedes if there was no such thing as the
> > commodore Amiga...... 1280x512 in 262,144 colours, I love the
> > Amiga .........
>
> Thats cool, have fun with it. And I'll stick with what I find suits me best
> and everyone is happy...'kay? Okay.
>
I aggree you seem to get on well with the arc (you've probably
got it set up better too), and it is definatly a nice machine,
(but it's too muched based on the BBC for my liking)
> Philip
>
> --
> Philip R. Banks Syntax: mail < ban...@kosmos.wcc.govt.nz > @@@@@/|
> Wesley suddenly underwent a gratutious total existence failure. "My God!@@@@/#|
> " cried Dr. Crusher, "What's happened to Wesley?". "Who cares?" replied @@@/##|
> Riker. "Agreed, #1," said Picard. "Mr. Data, take us out of here-Warp 9.@@/---|
> Engage." The Enterprise shot towards the stars. --- Minimalist Trek @/ |
Andreas Dehmel
>In article <2dj0h7$e...@gabriel.keele.ac.uk> cs...@seq1.keele.ac.uk (A.W. Oakden) writes:
> OS2 still takes more Ram than the machine comes with as standard!
>So what? These arguments that say "if it can't run in 2MB of memory
>and 10MB of hard disk space it's rubbish" are specious at best.
>If you want cheap, go buy a Commodore 64. I'm willing to buy the
>extra memory and disk space as required, just so long as the software
>is available to take advantage of it. If the software ain't there, it
>doesn't matter how much I stick in the machine, now does it? I'll take
>expensive options over no options at all.
> [...]
>ya. There are some 3 million OS/2 users now, so it's not like you exactly
>have a restricted market to sell into!
> Jordan
Funny thing, OS/2 (OS one-half...?). I know two guys who turned their
backs on Windoze in disgust to get OS/2 and ended up quite frustrated
because there simply wasn't ANY software for this OS. All right, you
can run Windoze software on it, but then you have the Windoze environment
which is just what you wanted to avoid. And if every once in a while
there IS a new program for OS/2 it costs even more than Windoze Software,
which is much too expensive in the first place.
There surely are a lot better OS on a PC than Windoze, but software
support is PRETTY low for any of those. How many programs are there
specifically tailored for usage with OS/2? Well, I got a 180 page product
directory for my Archimedes, which is probably hopelessly out of date
by now. Windoze won't sure as hell have any difficulties in topping that,
but from OS/2's point of view this is way out of reach.
So as far as I'm concerned OS/2 demands you to stick quite a bit into
your computer and end up with a good filing system and no software.
Bob Voisey
>Look, I just wish people would make sure they knew what they were talking about
>before they critise someone else's computer. I try to get most of my posts here
<grinning to himself knowing that Hugo was an Amiga fanatic long before he
got into the Archimedes>
Bob
PS. OK, so he's a bit rusty on the Amiga now.
--
Go away, else I shall taunt you a second time.
b...@cryton.demon.co.uk vox+44(749)670058 fax+44(749)670809 dat+44(749)670030
Mr M D Wooding
>|> In article <1993Dec2.1...@infodev.cam.ac.uk>,
>|> C.P. Brown <cpb...@hermes.cam.ac.uk> wrote:
>|> >All this talk about the ability to use Windows without a mouse making somehow a
>|> >better interface than RiscOS is frankly rather silly. Yes, Windows has lots of
>|> >keyboard shortcuts, but I find that most users tend to have to resort to them
>|> >becuase Windows mouse interface is akward and clumsy. Just some of the things I
>|> >really hate about Windows mouse interface:
I think being able to control everything from the keyboard is something to do
with US Government contracts -- if you can't do it with the keyboard, they don't
want to know... Yes, I know it's silly.
>presumably, when you click once, you let go of the button, right? The menu stays
>there... What if you dont want to select anything after all? You spend time
>moving the mouse to a neutral area of the screen, before clicking again..
>Hardly optimal.
Codswallop. Click again over the menu bar item and the menu disappears.
>What also irritates me is that Windows uses the select button
>for menu's. Now (from an Amiga point of view, because that's what I have),
>this is very annoying, because it means you can't make multiple selections
>from menus.
I think some apps may actually leave the menu up. It's a tiny percentage,
though.
It is very annoying that you have to keep opening the menu to make more selections.
>|> > - Floppy disk access seems to be more important than the ability to move the
>|> >mouse pointer.
>|>
>|> The Arc has a hardware driven pointer, and so cheats. Floppy accessing makes
>|> most computers chuggy until it's finished. Giving the pointer priority would
>|> make it take longer.
The pointer is only hardware driven in that it is not a part of the screen
as such, but a separate video entity. The OS still has to move it around in
the normal fashion. You can tell -- when you turn IRQs off the pointer stops
moving.
>Why should floppy disk access make the machine chuggy. When I access the floppy
>on my machine (again, this is from an Amiga perspective), it certainly doesn't
>seem chuggy, and all my programs continue to run with no noticible slowdown.
>(This problem on a PC basically comes down to a poorly designed architecture)
Yup.
>And do you have any comments about my other points? (no way to send a window to
>the back, and the active window must be at the front)
OS/2 suffers from these too, although I think IBM are doing something about it
(their `all windows active' idea -- daft name, I know).
Also note that most dialogue boxes freeze up the parent application, so you
can't e.g. move its window out the way. This is also very nasty.
--
(` t r a y l i g h t / This space \ PGP public key by email or
,_) (cs...@csv.warwick.ac.uk) / intentionally \ finger. Key fingerprint:
`That's the price you pay / left \ F8 43 53 BB AC 25 2A AD
for money' [Kill Me Ce Soir] / blank \ 17 0E 04 04 83 9E E5 7A
D.J.Davies
> >|> In article <JKH.93No...@whisker.lotus.ie>, j...@whisker.lotus.ie (Jordan K. Hubbard) writes:
> >|> >
> >|> > Um.. [clears throat] I'd kinda like an FPU, myself.. As soon as I can
> >|> > find somebody who'll sell me one for my A540, I'm buying one. Any
> >|> > suggestions? Everybody seems keener to offer upgrades for the A5000's
> >|> > these days. So much for being at the forefront. Sob..
>
> Forefront? What forefront? Acorn have dropped the A540, and in fact did so ages
> agoo. Sorry.
>
> >|> got one!), and while it can be classed as a co-pro, it does not come
> >|> under what the Amobea chappie was talking about.
>
> It *is* a processor. It's just that it only does FP instructions.
>
> >|> I seem to remember some Amobeas allowing multiple 680x0 processors (I.E.
> >|> CPU's) . Whilst I wouldn't mind a multi-arm700 m/c for Arch-BSD, I think
> >|> *ONE* ARM3 will do fine (at the minute).
Your missing the point of amiga coprocessors (there're not just
Maths copros and windows accelorators ( there are a total of 14
processors (NOT multiple 68xxx's) 64 bit block image
transfere's chip, 64 bit 16 mil colour co pro (Copper) etc.
Oh I'm the Ameoba Guy. /\/\//\/OO,
>
> Well, there's no reason the next Arc couldn't have multiprocessor support; I
> don't think the RISC OS programming interface would disallow it or anything.
>
> >I've never seen an Amiga with multiple 680x0's, apart from old 16 bit ones with a
> >32 bit CPU installed, keeping the old 68000 for backwards compatibility with bad
> >software. There was an april fools posting this year about a multiple 68040 and
> >060 Amiga, but that was just that, a joke.. There is a cute device called the
> >screamer from NewTek, which uworks in parallel with the video toaster and uses
> >several MIPS R4400's to speed up ray tracing...
>
> Yeah, and I'll bet it's not cheap, either...
Your right #10,000 plus /\/\/\/\/OO,
>
> >|> I honestly don't see what makes the amobeas so attractive... they have
> >|> a horrifying OS (or the Amiga 4000 I saw did), claims of high graphics
> >|> resolutions and BPP's, but which are all done in h/w using some sort
> >|> of cheating system which doesn't give True Colour (TM) and, most
> >|> importantly, they can't run !Axis or !Impress !!
>
> Well, the HAM modes are a bit of a cheat, but what do you usually want to
> display using all your millions of colours? Scanned photographic pictures and
> bits of artwork you've drawn yourself. Graduated fills are trivially easy, and
> you get a basic palette to draw your desktop in too. So I do object to the way
it's hailed as 262,144 colours when you couldn't actually display a random
> screen with all those colours on it, but when you come down to using it on a
> desktop, it does everything yer 256 colour mode does, and a bit more, so it can
> be useful.
>
> >Maybe you should try actually USING AmigaDOS, instead of just flaming it. It is
> >one of the most powerful single user OS'es in existance, cannot be matched for
> >multitasking, has lots of nice stuff such as datatypes, screens etc, and a really
> >powerful shell language (more powerful than a lot of unix stuff in a lot of
> >ways), that is also easy to use. As for graphics, 1400*592, 800*600 and 784*1100
> >in 262144 colours, or 256 REGISTERED colours from a 24 bit palette seem perfectly
> >adequate for base model graphics. If you want very high res 24 bit graphics, then
> >buy a GVP spectrum EGS card (1600*1280 in 256 colours, lower resoloutions in 15
> >and 24 bit), or some such. There is a lot of choice in this area. The AAA chipset
> >will be available sometime in 94, which gives 1280*1024 24 bit non-interlaced,
> >programable pixel clock, 8 channel 100KHz 16 bit sound, and lots of other nice
> >features. It's 64 bit and can support up to 16 megabytes of CHIP RAM, which can
> >be VRAM, DRAM, or a mixture of both.
>
> Well, since we're talking about future machines now, the next Arc will have a
> VIDC20 in it. As I understand it, it can support 4Mb of VRAM, therefore
> 1600x1200x16bpp and 1280x1024x24bpp. The VIDC20 is available NOW and will be
> fitted in new Acorn machines sometime next year. (OK, so it hasn't officially
> been announced, but they're not stupid enough not to, iyswim.)
>
> 16Mb of RAM. Ho hum, not much really; we've had the option since 1989 or so
> (A540); but no-one took it up 'cos most of the time you don't need it.
>
> John.
> --
> Disclaimer: I know only that I know nothing
D.J.Davies
> In article <1993Nov29.2...@infodev.cam.ac.uk>,
> R. Brown <rb1...@hermes.cam.ac.uk> wrote:
> >Seriously though, I think the point is well made. If I couldn't have an
> >Amiga, my
> >second choice of computer would be an Arch. Why would ANYONE want to use
> >Windoze?
> >:-)
>
> I propose a little experiment. I'll lock you in a room with a PC
> running MS Windows and an Archimedes, but I'll take away the mice from
> both machines. I'll come back after a week, and see which one you're
> using.
>
Actually I'd use neither!!!! (the Arc would'nt work!, and I
hate DrOS
> Ol
D.J.Davies
Ham & Ham 8 were design to cheeply display contineus tone
pictures at near 24bit quality without buying an expensive
display board and new monitor (you can also buy these if you
wish), I know 256 cols is'nt bad , but personally I'd like to
have the Choice of both!
/\/\/\/\/\/\/\/OO,
D.J.Davies
> In article <2dgjv4$i...@taki.nsict.org>, cl...@nsict.org (Clive Jones) writes:
> |> In article <1993Nov29.2...@infodev.cam.ac.uk> Robert Brown <rb1...@hermes.cam.ac.uk> writes:
> |> >Just thought I'd make the pint: The Amiga's the emulation specialist. Give us an
> |> >ARM CPU for Emplant (Or let the guys at UU have a go at emulating it. They seem
> |> >to think they can get an almost full speed 486 emulation out of a 68040...<grin>)
> |> >and a set of RISCOS ROMS, and we'll probably manage. Now, what would the Arch
> |> >need for an Amiga emulation?
> |>
> |> It's more plausible that an Amiga could be emulated on an Arc than
> |> vice-versa, I feel. Blitting is easier to do in software on an ARM
> |> than ARMcode is to do in a blitter. (-8
> |>
> |> I don't believe that a 486 could be emulated at more than about a
> |> tenth of full speed on a 68040.
> |>
> |> --Clive.
>
> Read my post. How much is an ARM chip these days? Plug one into the Emplant
> board, and use THAT to run ARMcode. The Amiga can do any necessary graphics
> processing, and we'll have RISCOS on a nice draggable screen, thankyou.
>
> the line? The Arch in it's present form CAN'T use all of the screen modes
> available to the Amiga, and emulating a lot of the other hardware would slow
> lot of programs are looking for a planar screen, and converting that will slow
> things down, especially if you're emulating a 68000 too (68020 probably, since
> this is now the base Amiga model). HAM modes might also be somewhat difficult.
> And the last time any of the people I know looked, you COULDN'T get the
> horizontal position of the display beam on current Archs, which you need to
> emulate the Copper at any plausible speed... OF course, if you're just going to
> emulate the OS, you could re-code the graphics.library (Nice having a modular OS
> isn't it (this IS serious. No flame (to the Arch) intended)), and tell it to
> refuse to open HAM screens, make all screens non-draggable, etc. but this would
> cause at least a good percentage of programs to refuse to run. Hardware level
> emulation would be A LOT more difficult, and would CRAWL. Ok, so plug in an Amiga
> chipset to help? Well, you've already covered most of the cost of an Amiga. Put a
> 68020 on the card too, and effectively it IS an Amiga (020s are REALLY cheap
> these days...). Oh yes, doing the disk drives could be fun, too.
>
> What I'm basically trying to say is that whilst it's obviously POSSIBLE to
> emulate an Amiga on an Arch, it's probably much easier/faster/cheaper to just buy
> an Amiga. However, it WOULD be possible to emulate an Arch on an Amiga.
>
> An OS level emulation in either direction would be easy. Hardware would be much
> more difficult. Modular operating systems make that sort of thing quite easy.
> (Actually, I've got an Arch owner sat next to me here who thinks he might have a
> go at Amiga emulation if I give him a hand. I may have to apologise for all of
> the above... :-)). Actually... anyone want to try the Arch on an Amiga? (Mind you,
> we're only talking OCS/1.3 emulation for the Arch, which is hardly sort of state
> of the art. A bit like emulating an Arch with an ARM2 and Arthur. Actually, why
> is it both Amiga and Arch owners don't like ANY desktop that's dark blue? :-) :-))
>
> This whole argument could change with the next release of the Amiga operating
> system (except perhaps the disks *grin*), which is apparently re-targettable in
> the graphics department.
>
> The thing about the 486 emulation came from an announcement from Utilities
> Unlimited, who're WRITING a PC emulator for their card. Apparently it was going
> to have a 486 in it, but they decided they didn't need it! (*I* thought it was
> unbelievable too, but it's not ME that's saying it...)
>
> Don't get me wrong. I *LIKE* the Arch. If the Amiga didn't exist, I'd probably
> have one. The two machines are probably the most 'original' platforms currently
> on the market! So why don't we all stop shouting at each other? :-)
I did'nt initially want to start a flaming session, but I quite
like the informed talk about technology from the Arch people. I
think that Robert has the right idea above in his last
statment. Both machines are very good it's windows I really
hate! (PC's)
>
> Robert Brown.
R. Brown
|> In article <1993Dec3.0...@infodev.cam.ac.uk> rb1...@hermes.cam.ac.uk writes:
|>
|> >Look, I just wish people would make sure they knew what they were talking about
|> >before they critise someone else's computer. I try to get most of my posts here
|>
|> <grinning to himself knowing that Hugo was an Amiga fanatic long before he
|> got into the Archimedes>
Well, he still got it wrong :-) (I must be one of the few Amiga owners who get
annoyed with the amount of rubbish posted by OTHER Amiga owners in assorted
.advocacy groups. Some of them are almost as bad as the PC owners :-) :-)).
|> Bob
|>
|> PS. OK, so he's a bit rusty on the Amiga now.
|> --
|> Go away, else I shall taunt you a second time.
|>
|> b...@cryton.demon.co.uk vox+44(749)670058 fax+44(749)670809 dat+44(749)670030
Robert Brown.
Adam Curtin
>Incidentally, you say:
>"The mouse is an integral part of RiscOS, and all systems come with one"
>
>What about the A4?
>
>considering it is a keyboard mouse simulator.
It isn't good, it's rubbish.
The bendable key on the SPARCbook is good.
A.
---
/home/adam/.sginature: No such file or directory
Hugo Fiennes
>|> Emulating 8 channels on a 4 channel sound chip takes all the amiga's cpu
>|> time - witness octamed on an A500 (total freeze during playing, not even
>|> the mousepointer moves - OK, so maybe this is partly by design, so you can't
>|> interfere, but it isn't doing much useful apart from playing music).
>
>OctaMED does not take all CPU time on an A500. It takes a noticable chunk of it,>
>but the mouse pointer moves quite freely. The probable reason it locked when
> you
>tried it is because OctaMED uses interrupts to play it's sound, and if you run
> it
>in 8 channel mode, on a machine with only CHIP RAM, in a high bus-contention
>graphics mode, then the machine spends all it's time servicing interrupts, and
>cannot run any other tasks. Your comment about not being able to interfere is
>silly. OctaMED allows you to edit scores whilst they are playing, as well as
>use most of the other features of the program. Try it on a mcahine with
> FASTRAM.
>You should see a noticable speedup.
I ran it on an A1000 with 1.5Mb ram (0.5Mb chip) with fastmemfirst as I
remember (it was quite a time ago).
>|> I believe the A1200 has hardware support for 8 channels, which would explain
>|> it.
>
>You believe wrong. The A1200 has the same 4 channel sound as the A500. The
> reason it runs faster is because the A1200 is a faster machine!
Ok - it was something I heard when the A1200 was released...
>|> (Octamed on an a500 sounds awful because you get max 14kHz sample rate per
>|> channel - yuck!).
>
>You should see it on an A1200. it sounds perfect, and takes up very little
> extra CPU time over playing 4 channels (somewhere in the region of 2%)
Well, on an A500/A1000 it sounds a long way from perfect - a very long way!
It's very obviously using a low sample rate.
Hugo Fiennes
>Look, I just wish people would make sure they knew what they were talking about
>before they critise someone else's computer. I try to get most of my posts here
>about the Arc read over by at least one Arc owner before I send them off, to
> make
>sure I haven't screwed up on what the Arc can/can't do. If other people would
>perhaps make an effort to be so careful, maybe less flame wars would develop...
Excuse me. I ran my BBS on an amiga for 2 years - an A1000. I've owned an
Amiga since 1985, and have programmed in C and assembler on it. My
experiences with OctaMED were on an A1000 with 1.5Mb ram (0.5Mb chip),
running as I remember in a 2-bit per pixel screen. The screen was not
updated, the sound was awful and the mouse pointer did not move. My system
had a 100Mb SCSI drive, and lots of other bits. I *do* know things about
the amiga, for example it has maximum 28kHz sample rate and attenuation
control for each sound channel. Ok, so I havn't seen the latest OctaMED,
but the one I used was awful. I don't have any A1200 owners handy to find
out wether it had an 8 channel sound chip, but seeing as it had the updated
video hardware it seemed possible.
Possibly, you should think before you flame.
Oliver Betts
Adam Curtin <ad...@ifeng.demon.co.uk> wrote:
>In article n...@news.mantis.co.uk, ol...@mantis.co.uk (Oliver Betts) writes:
>>What about the A4?
>>
>>Admittedly it has a keyboard mouse simulator, which is good -
>>considering it is a keyboard mouse simulator.
>
>It isn't good, it's rubbish.
>The bendable key on the SPARCbook is good.
That's as maybe, but it's a hardware solution, not a keyboard mouse
simulator.
Stephen Burke
> ya. There are some 3 million OS/2 users now, so it's not like you exactly
> have a restricted market to sell into!
Considering that this is only about double the number of sales of Acorn's BBC
MOS, and maybe a factor of 10 more than RISC OS, is this really a lot?!
--
e----><----p | Stephen Burke | Internet: bu...@vxdesy.desy.de
H H 1 | Gruppe FH1T (Lancaster) | DECnet: vxdesy::burke (13313::burke)
H H 11 | DESY, Notkestrasse 85 | BITNET: BURKE@DESYVAX or SB2@UKACRL
HHHHH 1 | 22603 Hamburg 52 | JANET: s...@uk.ac.rl.ib
H H 1 | Germany | Phone: +49 40 8998 2282
H H 11111 | HERA, the world's largest electron microscope!
D.Pead
>All this talk about the ability to use Windows without a mouse making somehow a
>better interface than RiscOS is frankly rather silly. Yes, Windows has lots of
Well, a 500 pound cheque tucked inside the manual wouldn't make Windows a
better operating system than RISC-OS. However, the fact that RISC-OS is
pretty well unusable without a mouse is a slight drawback. The two
areas of difficulty this causes are:
(a) Portable machines. If Acorn ever replace the A4 then maybe they should
consider the Mac Powerbook format (keyboard set back with wrist-pads and
a central trackball in front).
(b) Special Needs access. For educational purposes, providing hotkeys for
most essential functions (amongst other things) solves most problems, though.
Snag is, the things which make RISC-OS mouse dependent (the menu system,
the lack of the "current window" concept) happen to be my favorite features!
Bernard Jungen
In article <1993Dec3.0...@infodev.cam.ac.uk>, cpb...@cl.cam.ac.uk
(C.P. Brown) writes:
[stuff deleted]
|> You should see it on an A1200. it sounds perfect, and takes up very little extra
|> CPU time over playing 4 channels (somewhere in the region of 2%)
In Article <1993Dec3.0...@infodev.cam.ac.uk>, rb1...@cl.cam.ac.uk
(R. Brown) writes:
[stuff deleted]
|> Interesting. Since they use EXACTLY THE SAME CHIP, does the A500 have 8 channels
|> too? Serously, though, the A1200 has 4 channels, same as the 500. It just has
|> more processor time, and can therefore do 'HQ' mode with 8 channels.
Is this chip able to mix 8 channels to 4?
If not, playing 8 channels would require the processor to do a lot of number
crunching, so I guess extra CPU time is more than 2% (on the Arc it IS number
crunching).
If someone *really* knows Paula, please report!
Bernard.
Bernard Jungen (Member of BASS)
jun...@informatik.tu-muenchen.de (temp)
C.P. Brown
|> In article <1993Dec3.0...@infodev.cam.ac.uk> cpb...@cl.cam.ac.uk writes:
|>
|> >|> Emulating 8 channels on a 4 channel sound chip takes all the amiga's cpu
|> >|> time - witness octamed on an A500 (total freeze during playing, not even
|> >|> the mousepointer moves - OK, so maybe this is partly by design, so you can't
|> >|> interfere, but it isn't doing much useful apart from playing music).
|> >
|> >OctaMED does not take all CPU time on an A500. It takes a noticable chunk of it,>
|> >but the mouse pointer moves quite freely. The probable reason it locked when
|> > you
|> >tried it is because OctaMED uses interrupts to play it's sound, and if you run
|> > it
|> >in 8 channel mode, on a machine with only CHIP RAM, in a high bus-contention
|> >graphics mode, then the machine spends all it's time servicing interrupts, and
|> >cannot run any other tasks. Your comment about not being able to interfere is
|> >silly. OctaMED allows you to edit scores whilst they are playing, as well as
|> >use most of the other features of the program. Try it on a mcahine with
|> > FASTRAM.
|> >You should see a noticable speedup.
|>
|> I ran it on an A1000 with 1.5Mb ram (0.5Mb chip) with fastmemfirst as I
|> remember (it was quite a time ago).
Ah, fastmemfirst! So you must have been using AmigaDOS 1.x (yeuch!) and a very
old version of OctaMED? The earliest version I have used is V2 pro, and I can
assure you that it most certainly does not lock up, nor does it sound awful.
|>
|> >|> I believe the A1200 has hardware support for 8 channels, which would explain
|> >|> it.
|> >
|> >You believe wrong. The A1200 has the same 4 channel sound as the A500. The
|> > reason it runs faster is because the A1200 is a faster machine!
|>
|> Ok - it was something I heard when the A1200 was released...
AGA still has the same sound capibilities as vefore :-( The forthcomming AAA
chips (mid 94) apparently offer 8 chennel 16 but sound at 100KHz. just a tad
better than 4 channel 8 bit sound... :-)
|>
|> >|> (Octamed on an a500 sounds awful because you get max 14kHz sample rate per
|> >|> channel - yuck!).
|> >
|> >You should see it on an A1200. it sounds perfect, and takes up very little
|> > extra CPU time over playing 4 channels (somewhere in the region of 2%)
|>
|> Well, on an A500/A1000 it sounds a long way from perfect - a very long way!
|> It's very obviously using a low sample rate.
That's down to the module you were playing. I believe it aslo halvesd the sample
rates if it is running on a slower Amiga. (I've also not seen it running under
AmigaDOS 1.3, since I have used 3.0 for tha last year, and 2.04 for over a year
before that)
(BTW, if your last experience with AmigaDOS was 1.3, have a look at 3.0. You will
be very pleasantly surprised...)
R. Brown
|> In article <1993Dec3.0...@infodev.cam.ac.uk> rb1...@hermes.cam.ac.uk writes:
|>
|> >Look, I just wish people would make sure they knew what they were talking about
|> >before they critise someone else's computer. I try to get most of my posts here
|> >about the Arc read over by at least one Arc owner before I send them off, to
|> > make
|> >sure I haven't screwed up on what the Arc can/can't do. If other people would
|> >perhaps make an effort to be so careful, maybe less flame wars would develop...
|>
|> Excuse me. I ran my BBS on an amiga for 2 years - an A1000. I've owned an
|> Amiga since 1985, and have programmed in C and assembler on it. My
|> experiences with OctaMED were on an A1000 with 1.5Mb ram (0.5Mb chip),
|> running as I remember in a 2-bit per pixel screen. The screen was not
|> updated, the sound was awful and the mouse pointer did not move. My system
|> had a 100Mb SCSI drive, and lots of other bits. I *do* know things about
My system is plain vanilla A500 with fast RAM and GVP HD. The screen updates fine
in *3* bits per pixel, and the mouse pointer *DOES* move... I don't know WHAT you
did to your poor A1000... :-) (OctaMED default screen is 3 bpp, but there's an
option to turn it down to 2, which is recommended in the documentation for 68000
based Amigas...)
|> the amiga, for example it has maximum 28kHz sample rate and attenuation
Max 28kHz through DMA. Somewhat in excess of 50kHz if you use the processor,
even on a lowly 500 (just being silly and pedantic).
|> control for each sound channel. Ok, so I havn't seen the latest OctaMED,
|> but the one I used was awful. I don't have any A1200 owners handy to find
I've used assorted OctaMEDs since they came out (from magazine coverdisks,
though, but it's the same program...), and none have been as bad as you say.
Which version were you using? I'll look back and check it out...
|> out wether it had an 8 channel sound chip, but seeing as it had the updated
Didn't you see the flames of C= for NOT enhancing the sound? :-)
|> video hardware it seemed possible.
|>
|> Possibly, you should think before you flame.
Flaming? Well, maybe. See my earlier post about people over-advocating, though,
and note that I corrected AGAINST the Amiga with regard to the number of sound
channels. I like my Amiga as it is. I see no need to lie about it to try and make
other people see my point of view. If people think what I did was flaming, take a
look at comp.sys.(amiga|atari).advocacy sometime :-).
|> Hugo
|>
|> --
|> Hugo "I've suffered a fatal internal error and must exit immediately" Fiennes
|> alt...@cryton.demon.co.uk / vox +44 749 670058 / bbs +44 749 670030 2:252/102
|> ----------------------------- "...and that was my last treated digestive too"
Robert Brown.
Clive Jones
>DJDa...@cen.ex.ac.uk writes:
>> jdr...@cl.cam.ac.uk writes:
>> > cpb...@hermes.cam.ac.uk (C.P. Brown) writes:
>> > >In article <2d5jh4$s...@dunlop.cs.strath.ac.uk>, gpa...@cs.strath.ac.uk (Gary J Palmer C.S.3) writes:
>> > >|> In article <JKH.93No...@whisker.lotus.ie>, j...@whisker.lotus.ie (Jordan K. Hubbard) writes:
[ LOTS deleted ]
If there's one thing worse than someone who includes the entire text
of the previous article when following-up, it's someone who then goes
on to include the entire text of their article that included the
entire text of the previous article just to add one extra comment.
--Clive.
N W H Mailer
I have an A1200.
Eight channel sound works perfectly.
No real noticable reduction in quality.
Machine still far from sluggish.
And I don't know what it is about the Archimedes, but its sound is so terribly
tinny, even when amplified. Someone from EMR once explained the reason, but I've
forgotten. Can anyone enlighten?
R. Brown
I know Paula pretty well :-), and yes, it IS mixing using the processor. On an
A500 this IS a bit of a problem (40-50% of processor power gone, poof!), but on
the 1200 it's less than 5%, I think (based on measurements on a friend's...)
The mixing methods used vary depending on whether the 'HQ' mode is selected. I
know that without it's just a "Halve the sample value and add the results"
method, which effectively gives 7-bit sound at 28kHz (using the DMA to play the
result, so the processor isn't COMPLETELY crippled). Using the 'HQ' I'm not sure
what it does, but it does sounds better...
Anyway, this is getting a bit off the topic for this group. Anyone want to tell
me something about the Arc? That IS sort of why people read this group (yes, me
included!)
|>
|> Bernard.
|> Bernard Jungen (Member of BASS)
|> jun...@informatik.tu-muenchen.de (temp)
Robert Brown.
C.P. Brown
|> Is this chip able to mix 8 channels to 4?
|> If not, playing 8 channels would require the processor to do a lot of number
|> crunching, so I guess extra CPU time is more than 2% (on the Arc it IS number
|> crunching).
Paula's sound is 4 channel, 8 bit PCM stereo sound. It cannot do mixing of
samples. If you want to do that, the CPU has to do it all. However, this is
not a terribly intensive task (I don't know an awful lot aabout computer
generated sound, so forgive me if I'm wrong, but as far as I know, all one has
to do is take the average of the two samples at each point.) Bearing in mind that
the A1200 has a 68020 processor, and runs at up to 5 times the speed of an
original Amiga, multiplexing sound channels with the CPU is not vastyl processor
intensive.
|>
|> If someone *really* knows Paula, please report!
|>
|>
|> Bernard.
|> Bernard Jungen (Member of BASS)
|> jun...@informatik.tu-muenchen.de (temp)
--
R. Brown
[...]
|> If there's one thing worse than someone who includes the entire text
|> of the previous article when following-up, it's someone who then goes
|> on to include the entire text of their article that included the
|> entire text of the previous article just to add one extra comment.
|>
|> --Clive.
I agree. :-)
Robert Brown.
A G Jackson
|> ol...@mantis.co.uk writes:
|> > In article <1993Nov29.2...@infodev.cam.ac.uk>,
|> > R. Brown <rb1...@hermes.cam.ac.uk> wrote:
|> > >Seriously though, I think the point is well made. If I couldn't have an
|> > >Amiga, my
|> > >second choice of computer would be an Arch. Why would ANYONE want to use
|> > >Windoze?
|> > >:-)
|> >
|> > I propose a little experiment. I'll lock you in a room with a PC
|> > running MS Windows and an Archimedes, but I'll take away the mice from
|> > both machines. I'll come back after a week, and see which one you're
|> > using.
|> >
|>
|> Actually I'd use neither!!!! (the Arc would'nt work!, and I
|> hate DrOS
F12
*BASIC
[write a quick keyboard mouse emulator, or load in the one that you have on your
hard disc anyway]
QUIT
*Work.MouseEmul
*
Well, that would take the whole of a few minutes. How about putting an Amiga,
an Arc, a Mac and a PC in a room, taking away the keyboard, monitor and CPU
from all of them, and cutting the mouse cable in seven places? Would this
prove anything either?
Adrian
+--------------------------------------------------------------------------+
| Adrian Jackson | a.g.j...@durham.ac.uk | 27 North Bailey, DURHAM |
+--------------------------------------------------------------------------+
| - "So we're surrounded by absolutely nothing. There's a word for it. It's|
| what you get when there's nothing left and everything's been used up."|
| - "Yes. I think it's called the bill." |
| -- (Terry Pratchett, Eric) |
+--------------------------------------------------------------------------+
John Robinson
>In article <CT93008.93...@black.ox.ac.uk>,
>Robin Watts <ct9...@black.ox.ac.uk> wrote:
>>In article <2digr1$n...@news.mantis.co.uk> ol...@mantis.co.uk (Oliver Betts) writes:
>Are you implying that I'm stirring things up? Hmmm? Okay, maybe a little :)
Double, double, toil and trouble...
>>Keyboard shortcuts are great if they are standard, and only a
>>small set (Like C-z, C-x, C-c, C-v, f3, f4) but obscure ones
>>only slow you down.
No, they don't, unless you HAVE to use them (is that what you meant?).
>> To insist (as MS do) that *all* apps
>>must be drivable purely from the keyboard is insane. This
>>inevitably detracts from the usability of a system.
Does it? Why? I can't see why each menu option having a keyboard equivalent
stops the menus being well-structured and usable (just 'cos lots of Windoze'
apps have badly structured menus doesn't mean it's the fault of the keyboard
shortcuts).
>If you can't remember the shortcut, then you can press Alt and use the
>cursor keys to operate the menus. This is the major problem with this
>idea on RISC OS - this is fine for pull-down menus, but with pop-up ones
>it doesn't work - which menu do you bring up?
The one that the pointer's over. Just as when you press the middle mouse
button. And remember you can have KeyMouse loaded to do all this for you.
>BTW by pop-up, I don't mean menus on a button - I mean menus brought
>up by pressing the Menu button on a window. There's probably a better
>word, but it escapes me right now.
I think pop-up is the Acorn description for them anyway (but don't make me
swear to it, 'cos I won't).
>MS Windows is built on DOS, which is a programmers nightmare. Great - I
>have 32 Mb in my machine, but my C program can use 600k max, and to get
>anywhere near that you spend days fiddling with what gets loaded high.
>To get more I have to write XMS/EMS memory stuff, or compile with a DOS
>extender, which won't run under Windows. The major problem programming
>Windows itself (like I'm meant to be at the moment) is the size of the
>manuals. There are functions to do things no sane programmer would
>want, so you have to wade a lot to find what you want.
Last time I wrote on a PC I just compiled in HUGE mode and didn't worry about
it, but admittedly I didn't expect the program to work under Windoze.
>>[...]
>>Indeed, but of these shortcuts only C-x, C-r, and f3 are ever used
>>by users (Delete, Rename, Copy). Anything else it is faster to do by
>>the mouse.
>
>Depends how much you use the software - I can perform any common (to me)
>operation in !Zap using the keyboard, for eg. The only thing I can
>think of I use the menus for is where you need to (eg changing the
>indent or width). That's at least 40 shortcuts, I reckon.
That's not really true; !Zap is now (1.00) completely driveable with the
keyboard, if you so wish; any command can be done by doing (I seem to remember)
Ctrl-Escape then typing its name (eg INDENT or WIDTH) and also Ctrl-W lets you
set the width. Whether this is useful, I haven't yet decided; I shall have to
learn emacs (unfortunately) so I might start using the !Zap emacs keybindings,
but really, I'd rather stick with the keystrokes I've already learned for the
wonderful (?) editor E (wot we 'ave on this mainframe here, and also if desired
on the Unix systems).
David Thomas Richard Given
one day a week during which all the Acorn people and
Amiga people are forbidden to scream at each other,
and instead we can all scream at Apple, Microsoft
and IBM? That means we could all AGREE for a change!
Yours thinking-all-this-is-terribly-funny,
David Given dt...@st-andrews.ac.uk
John Robinson
>The Arc has a hardware driven pointer, and so cheats. Floppy accessing makes
>most computers chuggy until it's finished. Giving the pointer priority would
>make it take longer.
That's only half true; the Arc does still need the software to move the pointer
around the screen (but you're right in as much as it doesn't need to draw the
pointer in software).
One of the things I used to think the Arc did badly was disc i/o, in that the
machine grinds to a halt polling for it every time you want to do it. The Amiga
and Unices do it properly, blocking the process requesting the i/o and letting
everything else carry on OK, but the PC (with Windoze) is useless; start
ftp'ing something directly onto your floppy drive and it's ceased until the
floppy's finished, so perhaps I'm not so ashamed of the Arc after all. It could
be done properly on the Arc, except that you need to add preemptive
multitasking (so that if you start another process again, when the disc
operation completes, you can switch back to the process that requested the i/o)
and some folks might consider that this spoils some of the cleanness of RISC OS
(not to mention the fact that all filing systems with slow devices ie all but
RAM-based ones would need to be rewritten to support it, and controller cards
need interrupts also....)
John Robinson
>In article <2do4e8$c...@taki.nsict.org>, cl...@nsict.org (Clive Jones) writes:
>I agree. :-)
Thirded.
pj...@minster.york.ac.uk
>I don't mean to offend anyone here!
>
>But I don't like the Archimedes computer
>
>
>argargrhrhhghghhhh
>
>Shock Horror
>
>There is not much wrong with the machine only three things stop
>me from bying one instead of the wonderfull Amiga I have at the
>moment.......
>
>1) The Machine is too unstable (probably due to badly written
>software (see point two) ) I've been working with RiscOs 3.1
>A3000's for a year and a half now.
What?!? I've just re-read this and realised you're refering to
the Archimedes and not the Amiga! I don't believe it! An amiga
owner saying the Archimedes is unstable!?!?!
The amiga has got to be one of the most unstable systems I have
ever had the misfortune to use. Ok, it's got loads of colours
and may be fantastic for games, but as a computer for serious
use, it's terrible! I wouldn't dare use one for any important
work, as I just can't trust them not to crash. Hell, I just
switched one on the other day and it crashed.
>
>2) The badly written comercial software, The envoirnment is
>nice the chip is very fast (why no Co-Processors), of all the
>packages that I've used only three impressed me.
>
>3) The filosiphy (I can't spell to save my life) of the Arc is
>that software is modular If your wordprocessor has'ent got a
>table editor? then buy one. (What a pain in the Ass!)
Ah, this would involve running *two* packages on the desktop,
wouldn't it? At the same time! And transfer information between
them! And not crash! Blimey! I'll bet this sort of thing is a
bit foreign to you amiga owners, isn't it?
>
>Darren J Davies
>
>I would buy an archimedes if there was no such thing as the
>commodore Amiga...... 1280x512 in 262,144 colours, I love the
>Amiga .........
Do you really need all those colours to display a flashing red
box saying 'software error...' ? ;=)
>/\/\/\/\/\/\/\/\/OO,
>DJDa...@uk.ac.ex.cen
Yours putting-on-flame-proof-wellies-ly,
Paul
"These views are all my own work - please give generously"
Jordan K. Hubbard
>DJDa...@cen.ex.ac.uk writes:
>> jdr...@cl.cam.ac.uk writes:
>> > cpb...@hermes.cam.ac.uk (C.P. Brown) writes:
>> > >In article <2d5jh4$s...@dunlop.cs.strath.ac.uk>, gpa...@cs.strath.ac.uk (Gary J Palmer C.S.3) writes:
>> > >|> In article <JKH.93No...@whisker.lotus.ie>, j...@whisker.lotus.ie (Jordan K. Hubbard) writes:
[ LOTS deleted ]
If there's one thing worse than someone who includes the entire text
of the previous article when following-up, it's someone who then goes
on to include the entire text of their article that included the
HEY! Watch your attribution, Clive! You make it look like *I* was the one
including entire articles, when it was in fact Mr "DJDavies", who has also
been driving me insane with his two line postings that contain 200 lines
of > > > > > > > > > > followup drivel. My postings are entirely original
drivel! :-)
Jordan
--
(Jordan K. Hubbard) j...@violet.berkeley.edu, j...@al.org, j...@whisker.lotus.ie
Jordan K. Hubbard
I've just had a really good idea! Why not specify
one day a week during which all the Acorn people and
Amiga people are forbidden to scream at each other,
and instead we can all scream at Apple, Microsoft
and IBM? That means we could all AGREE for a change!
Here's an even better idea: How about 7 days a week where everyone
screams about acorn products and completely ignores Commodore, Apple,
Microsoft and IBM?
Advocate 1. [aedvekit] Anwalt m; Fuersprecher m; 2. [~keit] verteidigen,
befuerworten.
Erm.. Actually, I was going to post the dictionary definition of "Advocacy"
there in a snide attempt to illustrate how this group does almost nothing
to advocate acorn computers, chosing instead to trash the others, but
the only thing at hand was a German dictionary.. :-)
Kristjan Valur Jonsson
>>MS Windows is built on DOS, which is a programmers nightmare. Great - I
>>have 32 Mb in my machine, but my C program can use 600k max, and to get
>>anywhere near that you spend days fiddling with what gets loaded high.
>>To get more I have to write XMS/EMS memory stuff, or compile with a DOS
>>extender, which won't run under Windows. The major problem programming
>>Windows itself (like I'm meant to be at the moment) is the size of the
>>manuals. There are functions to do things no sane programmer would
>>want, so you have to wade a lot to find what you want.
>Last time I wrote on a PC I just compiled in HUGE mode and didn't worry about
>it, but admittedly I didn't expect the program to work under Windoze.
It still won't get you across the 640k barrier. Huge mode only eases some
of the pain induced by the 64k segments in 80x86 real mode.
In any case, Windows doesn't support programs written in huge mode, i.e.
programs with multiple data segments. It is possible to do it, but it rquires
an additional degree of complexity to your programs.
Kristjan
--
Kristjan Valur Jonsson | The individual does not qualify for
Student of mechanical engineering, | making decisions regarding the
University of Iceland | activities of the many.
k...@hengill.rhi.hi.is | (Helmut, 1993)
Philip R. Banks
> In article <JKH.93De...@whisker.lotus.ie>, j...@whisker.lotus.ie (Jordan K. Hubbard) writes:
> > ya. There are some 3 million OS/2 users now, so it's not like you exactly
> > have a restricted market to sell into!
>
> Considering that this is only about double the number of sales of Acorn's BBC
> MOS, and maybe a factor of 10 more than RISC OS, is this really a lot?!
Didn't Acorn pass the 1 million mark in selling Arcs a year or two back?
Even assuming only 60% of those machines run Risc OS (and thats a very
pessimistic estimate) this belies the 'factor of 10' bit...
Personally the figure is probably closer to a factor of 4 more OS/2 users.
Philip
--
Philip R. Banks Syntax: mail < ban...@kosmos.wcc.govt.nz > @@@@@/|
"Riker to Picard, I have both good news and bad news. The good news is @@@@/#|
that Wesley has opened that door at the shaft bottom for us by an adept @@@/##|
transformation of gravitational potential energy into kinetic energy. @@/---|
The bad news is he seems to have survived the drop..." ---ST:TWiSB XI @/ |
Kyle Dawkins
>bu...@vxdesy.desy.de (Stephen Burke) writes:
>
>> In article <JKH.93De...@whisker.lotus.ie>, j...@whisker.lotus.ie (Jordan K. Hubbard) writes:
>> > ya. There are some 3 million OS/2 users now, so it's not like you exactly
>> > have a restricted market to sell into!
>>
>> Considering that this is only about double the number of sales of Acorn's BBC
>> MOS, and maybe a factor of 10 more than RISC OS, is this really a lot?!
>
> Didn't Acorn pass the 1 million mark in selling Arcs a year or two back?
>Even assuming only 60% of those machines run Risc OS (and thats a very
>pessimistic estimate) this belies the 'factor of 10' bit...
>
> Personally the figure is probably closer to a factor of 4 more OS/2 users.
>
> Philip
Well, given that IBM just released OS/2 for Windows which is
fully functional and costs less than your average game and half
the price of your average application (Norton Desktop for Windows,
which makes Windows just bearable, is twice the price of OS/2 which
has a marvellous GUI and runs Windows to boot), and that
it seems to be selling pretty fast (went to buy it yesterday and the
two stores I made it to were totally sold out), I would estimate
that the "10" factor is perhaps not as far off as you might think.
Cheers.
Kyle.
daw...@music.mcgill.ca
D J Roberts
>In article <1993Dec3.1...@gps.leeds.ac.uk> eng3...@sun.leeds.ac.uk (N W H Mailer) writes:
>>
>>I have an A1200.
>>Eight channel sound works perfectly.
>>No real noticable reduction in quality.
>>Machine still far from sluggish.
>>And I don't know what it is about the Archimedes, but its sound is so terribly
>>tinny, even when amplified. Someone from EMR once explained the reason, but
>> I've
>>forgotten. Can anyone enlighten?
>
>speaker and audio jack socket), on the old models (I'm not sure about the new
>ones). If you remove them (as I have done), belive me, it sounds _a lot_
>better.
>
>As to why they are there in the first place??? I dunno
>
>--
>adq Email: term-time : a...@ed.ac.uk
> non-term-time: the...@arcade.demon.co.uk
>
>
The reason the arc has filters on its sound system (according to Acorn
Computing) is the fact that all computers produce high frequency
interference. The filters are there to stop that interference. It does
go on to say though that if you take them off you might be lucky and
not get any, you must be lucky!
Clive Jones
>In article <2do4e8$c...@taki.nsict.org> cl...@nsict.org (Clive Jones) writes:
> >DJDa...@cen.ex.ac.uk writes:
> >> jdr...@cl.cam.ac.uk writes:
> >> > cpb...@hermes.cam.ac.uk (C.P. Brown) writes:
> >> > >In article <2d5jh4$s...@dunlop.cs.strath.ac.uk>, gpa...@cs.strath.ac.uk (Gary J Palmer C.S.3) writes:
> >> > >|> In article <JKH.93No...@whisker.lotus.ie>, j...@whisker.lotus.ie (Jordan K. Hubbard) writes:
>
> [ LOTS deleted ]
>HEY! Watch your attribution, Clive! You make it look like *I* was the one
>including entire articles, when it was in fact Mr "DJDavies", who has also
>been driving me insane with his two line postings that contain 200 lines
>of > > > > > > > > > > followup drivel. My postings are entirely original
>drivel! :-)
Hmm. If I'd wanted to comment on your posting, then I'd have followed
it up, or at least have tweaked the posting I followed up to look as
if I'd started working from your article!
My intent was to show that the person in question had included
considerable parts of four previous postings.
--Clive.
Philip R. Banks
> Because the three I liked actually worked for a change.....
> without drive problems etc...
Drive problems? Oh are you meaning the lame disc protections and the like
often used? Yeah such are normally right royal pains to use and often get
removed fairly quickly...
> They are only lumbering on the PC I have a WP for the Amiga
> that is exactly like Ami Pro (with every function you can think
> of) (Final writer) and yet the prog only takes 500k and whole
> package including 120 postscript type 1 fonts and 100 colour
> eps images take 5 megs
That depends on what you define as 'every function you can think of'.
Personally most Arc software I use has all the functions I need, occasionally
they are lacking but I am a little specialised in what I use the Arc for. It
all is down to personal taste, some people love monolithic software, some
don't. *shrugs* Pick the system that suits ya best is my motto.
> I aggree you seem to get on well with the arc (you've probably
> got it set up better too), and it is definatly a nice machine,
> (but it's too muched based on the BBC for my liking)
Yes my machine is heavily customised in it's setup for my preference. And it
does what I want how I want with a few drawbacks. I can't ask for much more
than that. :) As I say if the Amiga does that for you, cool. I find flamewars
about which machine is better to be mostly assinine because it always get down
to whether the machine does what you want. And whatever machine does that for
you is the best machine *for you*.
Or it could simply be I get sick of marketroids telling me why their system
is best. :)
Philip
--
Philip R. Banks Syntax: mail < ban...@kosmos.wcc.govt.nz > @@@@@/|
You can rent your blues and photograph your soul, @@@@/#|
You can even dig some diamonds out of rock'n roll, @@@/##|
You can change the world, But if you lose control, @@/---|
Simon Glass
> I propose a little experiment. I'll lock you in a room with a PC
> running MS Windows and an Archimedes, but I'll take away the mice from
> both machines. I'll come back after a week, and see which one you're
> using.
Windoze users do it with underlines...
--
Simon
Sig line trunca
Simon Glass
> If you can't remember the shortcut, then you can press Alt and use the
> cursor keys to operate the menus. This is the major problem with this
> idea on RISC OS - this is fine for pull-down menus, but with pop-up ones
> it doesn't work - which menu do you bring up?
Oh yes that's a tough one :-) maybe the window with the caret...
> Research seems to show that the problem with pop-up menus is that they
> are less obvious. Indeed with a few freeware apps I've only discovered
> some menus after much use. The major advantage with them is that you
> don't need to move the mouse much to get menus up.
The first thing I do in a RISC OS is click menu to see what features are
available in each type of window (often only one). But I suppose if popup
menus were not standard then this would not be natural.
> indent or width). That's at least 40 shortcuts, I reckon.
Zap has hundreds if you include cursor keys, etc. But at least you can grow
into it.
Simon Glass
> I'm not in love with MS Windows (I much prefer my ARM3 Arc to this 66MHz
> 486DX2 I'm typing on now), but it does feel much more coherently
> designed than RISC OS. The most obvious example is that you can do
> almost everything without removing your hands from the keyboard.
> Various filer hacks and patches are around which add key shortcuts to
> the RISC OS filer, so users seem to want this.
Windoze designed? You have to be joking...how could anyone design such an
awful user interface, a system with handles instead of memory addresses, a
system which only allows 4096 malloc()s, that needs a reboot to change screen
resolution, a configuration system from hell...? Windoze congealed over three
versions in a desparate attempt to try and bring PCs kicking & screaming into
the UI world, and make MS a lot of money. There are lots of excellent books
about how it was developed, if you are looking for a good read.
The keyboard stuff, BTW, was because Gates was worried that without it no one
would buy windoze (2 I think, the one that didn't even allow windows to
overlap). At that time mice were considered a toy rather than something you
would find in an office. He decided this when the s/w was almost finished, so
they had to retofit keyboard support to windoze. The developers felt that
they could have done it a lot better if it were designed in from the start.
> Olly
> --
> Trained in Cambridge as an urban attack cyclist.
Training by foreign language students I suppose :-)
Simon Glass
> Amiga 1200, 6MB, 80 Meg harddisk, 25MHz 68882, 262144 colours on screen
You must have a very garish desktop. Mine has only 13-14 at the moment...or
do you have a backdrop picture?
Oliver Betts
John Robinson <jdr...@phx.cam.ac.uk> wrote:
>One of the things I used to think the Arc did badly was disc i/o, in that the
>machine grinds to a halt polling for it every time you want to do it. The Amiga
>and Unices do it properly, blocking the process requesting the i/o and letting
>everything else carry on OK, but the PC (with Windoze) is useless; start
>ftp'ing something directly onto your floppy drive and it's ceased until the
>floppy's finished, so perhaps I'm not so ashamed of the Arc after all. It could
As I type this reply into a telnet session under Windows, I'm ftp-ing
direct to a floppy (from a local machine, so the ftp is flinging data at
quite a rate). The software doing the ftp is WinQVT/Net v3.8. Typing
is a bit ponderous, but the machine is hardly useless. If I try playing
minesweeper, it's a bit more responsive than playing !MineHunt while
copying to floppy on an Arc (to give you some idea).
Hardly "ceased" at any rate.
Ol
Oliver Betts
D.Pead <rs...@unicorn.nott.ac.uk> wrote:
>(a) Portable machines. If Acorn ever replace the A4 then maybe they should
>consider the Mac Powerbook format (keyboard set back with wrist-pads and
>a central trackball in front).
Unfortunately, all the existing "nice" solutions to the problem will get
you sued into the ground if you try them. You're very unlikely to see
the powerbook's arangement (which *is* very usuable) except on a
powerbook.
Simon Glass
> >commodore Amiga...... 1280x512 in 262,144 colours, I love the
> >Amiga .........
I hope that's not your only reason...the VIDC20 will do 1280x512 (or more
usefully on a 20" monitor, 1600x600) in 32k colours. Will you be buying an
Arc next year...I think not.
Oliver Betts
Simon Glass <sgl...@sglass.demon.co.uk> wrote:
>In article <2dlca2$1...@news.mantis.co.uk> ol...@mantis.co.uk (Oliver Betts) writes:
>> If you can't remember the shortcut, then you can press Alt and use the
>> cursor keys to operate the menus. This is the major problem with this
>> idea on RISC OS - this is fine for pull-down menus, but with pop-up ones
>> it doesn't work - which menu do you bring up?
>
>Oh yes that's a tough one :-) maybe the window with the caret...
Except many windows never gain the caret - filer windows are the obvious
example on vanilla RISC OS. You may also get different menus for
different areas of the window (eg a menu of filetypes over a filetype
writable icon). You could use the pointer's position, but then you need
to be able to move the pointer by key presses.
>> indent or width). That's at least 40 shortcuts, I reckon.
>
>Zap has hundreds if you include cursor keys, etc. But at least you can grow
>into it.
I know. The point I was answering was that most apps have lots of key
short-cuts, but users only use half a dozen. This is untrue (for me at
least) on apps that get used a lot.