AGA demos
s.m.windmill
It would help if Commodore got *their* finger out, and released the tech. info
for the AGA chipset.
Don't blame the coders for Commodore's stupidity.
Sayonara!
Si.
--
Si Windmill - s.m.wi...@bradford.ac.uk
"Hey, hey, hey! Don't come near _me_ with that tub of cold water!"
J.P. Hillenburg
>
> It would help if Commodore got *their* finger out, and released the tech. info
> for the AGA chipset.
>
> Don't blame the coders for Commodore's stupidity.
>
> Sayonara!
>
> Si.
Bah. The Amiga would have been much farther along today if it
weren't for idiot bang-on-the-hardware types. With an 030/etc,
you don't NEED to bang bang bang. Responsible programmers don't
bypass an OS which ultimately benefits them.
Chris Green has shown that impressive graphics can be done
WITH the OS. Learn a lesson from the wise, bud.
--
Joseph Hillenburg (xterm), NPS Technologies, Bloomington, Indiana
UUCP: j...@anaconda.Bloomington.IN.US \ /
"My OTHER computer is a CM-5" +----o0o----+
Long live the Hornet
Ranma
: It would help if Commodore got *their* finger out, and released the tech. info
: for the AGA chipset.
: Don't blame the coders for Commodore's stupidity.
Hey, I'm not saying that good software cannot be written OS legally - but in
Europe at least, the Amiga has been built by 'metal-bashers' - these people
have been responsible for some of the most ingenious code around.
Of course - there is a big difference between low level coding and just
poor coding. When people ignore any guidelines set by C= about future
changes, thereby ensuring that their demo is incompatible with future
hardware - now that is annoying.
All I'm asking for is for C= to release an updated HRM. Please?
Hans Guijt
>From: j...@anaconda.Bloomington.IN.US (J.P. Hillenburg)
>Subject: Re: AGA demos
>Date: 8 Nov 93 12:11:47 EST
>In article <1993Nov8....@bradford.ac.uk> S.M.Wi...@bradford.ac.uk (s.m.windmill) writes:
>>
>> It would help if Commodore got *their* finger out, and released the tech. info
>> for the AGA chipset.
>>
>> Don't blame the coders for Commodore's stupidity.
>>
>> Sayonara!
>>
>> Si.
>
>Bah. The Amiga would have been much farther along today if it
>weren't for idiot bang-on-the-hardware types. With an 030/etc,
>you don't NEED to bang bang bang. Responsible programmers don't
>bypass an OS which ultimately benefits them.
So, we all buy 4000's, so we can do the same or less with the OS that we
could do with hardware banging on the a500? That's spending money for
nothing.
>Chris Green has shown that impressive graphics can be done
>WITH the OS. Learn a lesson from the wise, bud.
The wise? I bet he writes directly into screen memory. That is poking a
system structure, which is illegal (according to the manuals).
Oh, and one other point: why should YOU decide if we are allowed to bang the
hardware?
Hans
Martin Blom
>It would help if Commodore got *their* finger out, and released the tech. info
>for the AGA chipset.
>Don't blame the coders for Commodore's stupidity.
All scene-coders have the AGA reference manual.
(s.d.s!)
>Sayonara!
>Si.
>--
>Si Windmill - s.m.wi...@bradford.ac.uk
>"Hey, hey, hey! Don't come near _me_ with that tub of cold water!"
>
--
+---------------------------+ Ahh.. btw: ///
| Martin Blom LiTH Sweden | ///
| EMail: l...@lysator.liu.se | \\\///
+---------------------------+ \XX/ rules.
Jyrki Saarinen
>It would help if Commodore got *their* finger out, and released the tech. info
>for the AGA chipset.
>Don't blame the coders for Commodore's stupidity.
AGA Hardware Reference manual is released, but only for game companies..
But, every good coder has it as an file.
* Offline Orbit 0.22 (beta) *
J.P. Hillenburg
> In article <jph....@anaconda.Bloomington.IN.US> j...@anaconda.Bloomington.IN.US (J.P. Hillenburg) writes:
> >From: j...@anaconda.Bloomington.IN.US (J.P. Hillenburg)
> >weren't for idiot bang-on-the-hardware types. With an 030/etc,
> >you don't NEED to bang bang bang. Responsible programmers don't
> >bypass an OS which ultimately benefits them.
>
> So, we all buy 4000's, so we can do the same or less with the OS that we
> could do with hardware banging on the a500? That's spending money for
> nothing.
The A500 is long since obsolete. And Green's demo (as WELL as others)
shows that you CAN do impressive AGA graphics through the OS.
> >Chris Green has shown that impressive graphics can be done
> >WITH the OS. Learn a lesson from the wise, bud.
>
> The wise? I bet he writes directly into screen memory. That is poking a
> system structure, which is illegal (according to the manuals).
No, Chris Green is a C= Software Engineer. He does the entire demo
through the OS. If I remember correctly, it even multitasks. (But
I am not for sure about that, not having an AGA machine myself...)
> Oh, and one other point: why should YOU decide if we are allowed to bang the
> hardware?
I am not, but C= has the right to release the specs or not, and if
they decide that releasing specs would drag down the machine, they
have that choice.
M.P. Francis
Is that the Chris Green demo which flickers when you move the mouse pointer
and jerks every few seconds? I don't want to start the hardware vs OS debate
*again*, but it isn't possible to achieve the sort of magic you see in demos
by using OS functions, end of story.
ok, you can go some way to doing a demo with the OS, but it won't be anything
near as good as one which has full control over its faculties. don't get me
wrong; banging the hardware with anything other than demos and games (both
having short life-spans) is plain wrong and stupid. but demos and (action)
games need to get as close to the hardware as possible (I believe the CD^32
has a good solution to this).
Matt Francis, University of Newcastle-upon-Tyne
(m.p.f...@newcastle.ac.uk)
Ralph Barbagallo
to be able to run the demo off of a hard drive from WB and return to the
WB when the demo is finished. That's all I ask...
--
Ralph A. Barbagallo III --- rbar...@cs.uml.edu --- Only AMIGA makes it
Possible...
Only C O M M O D O R E stands in the way....
ZzzzzZZZzzzzzzZzzzzzzz.... .
J.P. Hillenburg
>
> but demos and (action)
> games need to get as close to the hardware as possible (I believe the CD^32
> has a good solution to this).
specialfx.library, which lets you do low-level tricks in
a compatible manner.
Nils Liaaen Corneliusen
In article <1993Nov8....@bradford.ac.uk>, S.M.Wi...@bradford.ac.uk (s.m.windmill) writes:
>
> It would help if Commodore got *their* finger out, and released the tech. info
> for the AGA chipset.
>
> Don't blame the coders for Commodore's stupidity.
Excuse me, but that's *not* stupid. The AAA specs states that it will not be AGA
hw level compatible, only ECS. Take AGA for what it is, an ECS hack to allow up to
8 bpls and a 24 bit palette.
------------------------------------------------------------
Nils Corneliusen Triumph Software
net: nil...@ifi.uio.no Amiga 4000/040
phone: +47 22 26 34 05 "Greed is good. Greed works."
SM WINDMILL
: It would help if Commodore got *their* finger out, and released the tech. info
: for the AGA chipset.
: Don't blame the coders for Commodore's stupidity.
Ah, but I never said C= were stupid just for not releasing the info, did I?
I was merely bemoaning the fact that they should sort out just what
'standards' they are trying to promote. Let's face it, properly marketed,
with a decent hardware base (not just updating 80s hardware), the Amiga
would be ruling the world of computers, not just the world of hobbyists...
Hans Guijt
>>weren't for idiot bang-on-the-hardware types. With an 030/etc,
>>you don't NEED to bang bang bang. Responsible programmers don't
>>bypass an OS which ultimately benefits them.
Look, demos are around for a month before they are forgotten. What's wrong
about banging the hardware? Afraid your demo won't run on the next
generation of amiga's? Hahahaha!
And once again, I didn't buy a 68030 so that I could do the same legally
that I could do illegally before. I want to harness the full power of the
machine, and that includes using copper, blitter, and videochips, parts of
which are inaccessible through the OS.
But, all this implies that I don't want to do anything but bang the
hardware. WRONG! Some time ago I posted this list:
1. Applications should always use the OS for everything. Even better, they
should all follow the styleguide as well. This allows for some real
productivity on your machine.
2. Games should use the OS where possible, for things like loading data (
allows for HD installability), allocating memory, etc., but are allowed to
take over the machine when they need to in order to enhance performance.
3. Demos can do what they want. Let them bang the hardware - noone pays for
demos, noone is unhappy when demos fail (or at least not for long), noone
remembers demos one month after release.
4. AMOS should be burned (I just saw mini-office...).
I believe this is fairly reasonable.
>>Chris Green has shown that impressive graphics can be done
>>WITH the OS. Learn a lesson from the wise, bud.
>
>Is that the Chris Green demo which flickers when you move the mouse pointer
>and jerks every few seconds? I don't want to start the hardware vs OS debate
>*again*, but it isn't possible to achieve the sort of magic you see in demos
>by using OS functions, end of story.
Hear, hear!
>ok, you can go some way to doing a demo with the OS, but it won't be anything
>near as good as one which has full control over its faculties. don't get me
>wrong; banging the hardware with anything other than demos and games (both
>having short life-spans) is plain wrong and stupid. but demos and (action)
>games need to get as close to the hardware as possible (I believe the CD^32
>has a good solution to this).
As I said.
Hans
SM WINDMILL
OS vs hardware war!
Remember, this *is* alt.sys.amiga.DEMOS - and as the last poster said
(more or less) "Who cares if your demo won't run on some new super
configuration?". Demo compatibility should only be thought of as an added
bonus, not a requirement.
Of course, programmers of *any* commercial software have a responsibility
to ensure future compatibility of software that people have gone out and
bought, but once again, if C= gave out proper guidelines about very low
level programming, you could hardware-hit AND have compatibility.
Oh, and thanks to those people who have sent me tech info on
AGA/AA/whatever Commodore decide to call it next week....
Sayonara,
Morten Eriksen
|> > In article <jph....@anaconda.Bloomington.IN.US> j...@anaconda.Bloomington.IN.US (J.P. Hillenburg) writes:
|> > >From: j...@anaconda.Bloomington.IN.US (J.P. Hillenburg)
|> shows that you CAN do impressive AGA graphics through the OS.
So what? The point in making demos is not to make impressive graphics routines,
but to make the _most_ impressive graphics routines possible. And your graphics
routines can be made even more impressive if you bang the hardware directly,
instead of going through the OS.
Morten
Carl Henrik Aaby
In article <2bt0s6$6...@sognsvann.ifi.uio.no>, nil...@ifi.uio.no (Nils Liaaen Corneliusen) writes:
>
> In article <1993Nov8....@bradford.ac.uk>, S.M.Wi...@bradford.ac.uk (s.m.windmill) writes:
> >
> > It would help if Commodore got *their* finger out, and released the tech. info
> > for the AGA chipset.
> >
> > Don't blame the coders for Commodore's stupidity.
>
> Excuse me, but that's *not* stupid. The AAA specs states that it will not be AGA
> hw level compatible, only ECS. Take AGA for what it is, an ECS hack to allow up to
> 8 bpls and a 24 bit palette.
Who said that ECS wasn't an OCS hack???
>
>
> ------------------------------------------------------------
> Nils Corneliusen Triumph Software
> net: nil...@ifi.uio.no Amiga 4000/040
> phone: +47 22 26 34 05 "Greed is good. Greed works."
------------------------------------------------------------
INCLUDE "disclaimer.i" Amiga 4000/030
Carl H Aaby Smeagol/Triumph
internet: ca...@ifi.uio.no "Sex is fun, said the nun"
Nils Liaaen Corneliusen
In article <2c09gm$5...@gymir.ifi.uio.no>, ca...@ifi.uio.no (Carl Henrik Aaby) writes:
>
> In article <2bt0s6$6...@sognsvann.ifi.uio.no>, nil...@ifi.uio.no (Nils Liaaen Corneliusen) writes:
> >
> > Excuse me, but that's *not* stupid. The AAA specs states that it will not be AGA
> > hw level compatible, only ECS. Take AGA for what it is, an ECS hack to allow up to
> > 8 bpls and a 24 bit palette.
>
> Who said that ECS wasn't an OCS hack???
>
There's hacks and there's hacks :-). ECS was pretty ok since the added hw regs and
features were quite few. I guess we can just call it an improvement, not a "new
and revolutionary" chipset.
AGA on the other hand adda 8 bpls and 24 bit palette into a chipset that
never was intended for for more than 5 and 4 of the same things. Quite a mean feat,
quite easy to do, but nothing you can ever add more to.
Hans Guijt
personally, I'd like to send my opinion into the big, bad world. Read on:
>Or unless they do not care that the demo's or software do not run on
>machines in the future. I have looked at alot of AGA games and Demos just
>to find out they do not run on my 4000. I have also found previous demos
>that will not run on anythyng but the 68000. If done correctly you do not
>have to bang the hardware and still get great results. I just wish that
>alot of people will learn that in the near future. But ofcourse
>after doing it for 5 years people get used to it.
I think you put too much emphasis on 'banging the hardware' for your
incompatibility problems.
I have a 4000 too, and I did run into trouble. But not all is banging
the hardware, in fact banging the hardware is the least of your
problems! The real trouble arises when people do the following:
- Use 24-bit pointer (what do you mean the 4000 has 32-bit address
space?)
- Reference chip-registers on a ghost-address. The chip-registers
were mirrorred on certain locations on the a500. They no longer are
on the 4000, but hey, using a ghost-address is cool, sorry k00l!
- Assume memory exists someplace specific (like $c00000). Simply
untrue.
- Assume a program is always loaded someplace specific. Again not
true.
- Doing undocumented things with the OS, like jumping directly into
the $fc0000 range (this *IS* a big problem, thanks to Abacus...)
- Assuming that your processor is a certain speed, allowing for delay-
loops (my '30 eats delay loops...)
- Using self-modifying code (yum-yum says the cache).
I can probably think of 20 others, but this should suffice. The funny
thing is: hardware banging, if done by the rules set down in the
hardware reference manual, almost never fails!
Oh, and one other thing, there are no official 3.0 docs out, nor is
there any information on the AGA chips. This means that when
something is written specifically for AGA (ie. uses more colors on a
screen than ECS could handle, or uses data-types, etc.), that piece
of software is unreliable. You should avoid using AGA software on
your AGA machine.
As for those 'great results' you mention: this is simply untrue, and
I wish people would understand. The OS cannot keep up with custom-
written routines.
Hans
Andrew Zendic
JH> In article <1993Nov8....@bradford.ac.uk>
JH> S.M.Wi...@bradford.ac.uk (s.m.windmill) writes:
JH> >
JH> > It would help if Commodore got *their* finger out, and released the
JH> tech. info
JH> > for the AGA chipset.
JH> >
JH> > Don't blame the coders for Commodore's stupidity.
JH> >
JH> > Sayonara!
JH> >
JH> > Si.
JH> Bah. The Amiga would have been much farther along today if it
JH> weren't for idiot bang-on-the-hardware types. With an 030/etc,
JH> you don't NEED to bang bang bang. Responsible programmers don't
JH> bypass an OS which ultimately benefits them.
SINCE WHEN have demo coders been responsible programmers? The
objective, in most cases, of not only banging the hardware but of creating a
demo in the first place is to be 1> creative 2> rebellious 3> semi-legal.
JH> Chris Green has shown that impressive graphics can be done
JH> WITH the OS. Learn a lesson from the wise, bud.
Disney has shown that impressive graphics can come out of a Silicon
Grpahics machine. Get 'em how you can get 'em, I say. Bang hardware if ya
wanna, if you don't, don't bang others who do. If you want to program it
legally, do so.
--
/\/\ L...@Tezcat.Chi.IL.US ||| Face/MJD/Lex/SGT/Everyone Else
\/|> Andrew Zendic {0} -Youth is wasted on the Young. -G.B.Shaw
All opinions expressed are my own, and not neccesarily those of the host
system. It is further my opinion that Barney should be destroyed.