M2 tidbits
BEwert
A few tidbits of info from my latest Electronic Engineering Times. They
have an article on the 3DO/Samsung joint venture (pp29-30).
"The real strength of 3DO lies in multimedia architecture/design talent" a
Hugh Martin quote.
"Over the past several years, 3DO engineers have significantly advanced
hardware system designs and software drivers, in addition to developing
software modems, MPEG codecs, and encryption systems. Much of that
knowledge was stitched into the ASIC at the heart of the M2 system-the BDA
chip, short for Bulldog ASIC, derived from the original project name. The
BDA integrates an audio DSP as well as an MPEG1 video decoder--soon to be
upgraded to MPEG2 for DVD-compatible systems--and advanced audio/video
graphics capability with a 3D graphics engine capable of offering 500,000
triangles per second, in addition to a separate 3D setup engine to free
the processing power of the two PowerPC's. The ASIC is also capable of
decoding multiple video streams simultaeneously, while its video stream
can be mapped onto an object as texture. " [Note MPEG2/DVD connection
confirmed. Software modem is a new thing.]
"As far as 3DO's 128-bit system development is concerned, Toby Farrand
said, "We're very far along."
Interesting...
E @ T&T
http://www.altacam.com
Leo L. Schwab
In article <5ee744$f...@lana.zippo.com>, <Nobody's_Perfect> wrote:
>> in addition to a separate 3D setup engine to free
>>the processing power of the two PowerPC's.
>
You know, I think you need to re-read a chapter or two in your
self-help book. Affirmations only work for internal beliefs, not
empirical truths.
The setup engine is pipelined directly into the triangle engine.
There is no "nightmare," it works perfectly seamlessly.
> No wonder
> why developers are calling M2 the most difficult console to develop for,
> ever.
>
Citations, please. Names, affiliations, email addresses, phone
numbers. **WHO** are these fabled developers who are having so much
trouble? My colleagues and I can help them.
Let me be abundantly clear: I have made and am making an open and
standing offer to anyone and everyone developing software for M2. If you
are developing for M2 in any capacity at all, and are having trouble, get
in contact with me. My colleagues and I will do our level best to help.
> Even N64 renders 600,000 t-mapped triangles/s. Only when you turn
> on every feature does N64's poly count goes down to 160,000
> trianlges/s. This verifies Acclaim and Capcom's claims that M2 is not
> better than N64.
>
*sigh* And just who made *this* claim? Until you can provide me
and the readers of this group some concrete, verifiable sources for this
stuff, I'm going to have to assume that you are *making all this up.*
Either that or someone is deliberately feeding you bogus data, probably
for the purpose of amusing themselves at how gullible you are.
M2 can do 500K with all the features turned on. How is this "not
much better" than N64's 160K?
>>"As far as 3DO's 128-bit system development is concerned, Toby Farrand
>>said, "We're very far along."
>
> M3 is dead.
>
Actually, it's called MX, and I fear you could be right on this
point. MX has no customer (last I heard) and, IMHO, is unlikely to get one.
_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_
Leo L. Schwab -- Digital Spellweaver ew...@best.com ..or..
\_ -_ http://www.best.com/~ewhac/ ew...@well.com
O----^o Recumbent Bikes: The Only Way To Fly. (pronounced "EH-wack")
Hiroshima '45. Chernobyl '86. Windows '95.
Todd Allendorf
Nobody's_Perfect wrote:
>
I'm only responding to this troll because there might be a few
individuals out there that don't know Nobody (AKA Clueless_Coward) has
absolutely no knowledge about anything he is talking about.
> In article <19970218170...@ladder01.news.aol.com>, bew...@aol.com says...
> >
> >"The real strength of 3DO lies in multimedia architecture/design talent" a
> >Hugh Martin quote.
> >"Over the past several years, 3DO engineers have significantly advanced
> >hardware system designs and software drivers, in addition to developing
> >software modems, MPEG codecs, and encryption systems.
>
> Well, then why don't we see the commercial products based on 3DO
> techonology? I have never seen any of products you mentioned above
> bearing 3DO name.
>
YOU don't see them because $10,000 MPEG encoding systems aren't sold at
the Kaybee Toy Stores that you frequent.
> > Much of that
> >knowledge was stitched into the ASIC at the heart of the M2 system-the BDA
> >chip, short for Bulldog ASIC, derived from the original project name. The
> >BDA integrates an audio DSP as well as an MPEG1 video decoder--soon to be
> >upgraded to MPEG2 for DVD-compatible systems--and advanced audio/video
> >graphics capability with a 3D graphics engine capable of offering 500,000
> >triangles per second,
>
> That is not very good. I know of a $50 Phillips chip that renders
> 1 million triangles/s with t-mapping and G-shading.
>
1) You don't know how much the BDA costs. 2) The "Phillips chip"
doesn't do MPEG or audio. 3) How many mass market consumer products
are/will be made with this "Phillips chip" 4) MEI didn't license
"Phillips chip" for $100 million dollars. 5) Who cares?
> Even N64 renders 600,000 t-mapped triangles/s. Only when you turn
> on every feature does N64's poly count goes down to 160,000
> trianlges/s. This verifies Acclaim and Capcom's claims that M2 is not
> better than N64.
>
Oooh. Let's see some hard source for that 600,000 number. Under the
same non-game, hype, spec-sheet criteria that you could use (i.e. no
lighting, one continuous strip, less than 20 pixel polygons, all on the
same scanline, no filtering) to MAYBE get that 600,000 on the N64, the
M2 can do 1 million/textured triangles per second. Why doesn't Hugh say
1M tris/sec? Simple. He doesn't want to quote theoretical spec numbers
he knows can't be achieved in a real world games. Remember how silly
Sony and Sega looked with their "500,000 tris/sec claims" when
eventually it became clear no game would ever hit those nubmers? While
the N64 drops to about 150,000 (being generous with SGIs own numbers),
the M2 in a real-world application can get about 480,000 triangles/sec.
The test program developed by Mark Cerny of Universal Interactive (among
his credits are Marble Madness, Sonic 2, Crash&Burn, Disruptor, and
Crash Bandicoot) has about a hundred objects randomly moving about the
screen, multiple textures, three light sources (one directional, one
point, one spotlight), fogging, and tri-linear mip-mapping. The test
program (which every developer receives) is a C-program specifically
designed to simulate the performance of a real-world game. It has
multiple textures being loaded, objects continually updating their
transform matrices, and processing of game logic.
As I have told you in the past, Acclaim never had an M2 dev system (and
the bozo you've been quoting never even saw one) and your Capcom
"source" has never been named or confirmed. BTW, Capcom also gave 3DO a
benchmark program that it used on the N64. With a single processor,
the M2 was 60% faster than the N64. On a two processor system, the
benchmark showed M2 more than twice as fast the N64.
BTW, being the coward that you are, you never accepted my challenge that
you would leave all the videogame groups forever if an M2-based arcade
system from either Capcom or Konami ever shipped. Still chicken?
> > in addition to a separate 3D setup engine to free
> >the processing power of the two PowerPC's.
>
> Wow, what a messy architecture. Coder's worst nightmare. No wonder
> why developers are calling M2 the most difficult console to develop for,
> ever.
>
If you even knew what a setup engine was, you would know it is
transparent to programmers. Please quote even ONE M2 developer (that
would be a developer that has actually used an M2 dev system) that has
said the M2 is the most difficult console to develop for. Every
developer I've talked to has said the opposite. Its the easiest console
to develop for. The reasons are 1) A unified memory architecture. No
more shuffling data from memory pool to memory pool crap (like on the
PSX or Saturn) which makes resource management a nightmare and kills
performance. 2) 8MB of memory. Plenty of room for a competent
developer. No more shoe-horning games into < 4MB (N64, Saturn, PSX) 3)
Clean multi-processor design supported by a small, efficient,
multi-tasking OS. No more timing or interrupt nightmares that occur on
the Saturn. Utilizing both processors is trivial. 4) Efficient
libraries for handling multiple, overlapping data streams from the CD.
No need for developers to keep re-inventing the wheel like they have to
on the PSX and Saturn. Ever wonder why the 3DO Opera version of WCIII
has transparent loads while the PC and PSX versions have long pauses?
3DO's preemptive multi-tasking OS and streaming libraries make it
possible.
To quote Alien V. Predator's Andrew Whittaker (working on Power Crystal)
"M2 really is the single finest piece of hardware we've worked with. Its
power will hit the world of entertainment software like a tidal wave."
> >Interesting...
>
> M2 is too little, too late. It is not going to be able to match
> upcoming Lockheed's machine for Sega(this Lockheed machine supposely
> exceeds Model3's performence). I think Sony canned original PSX2 design
> because of this Lockheed Martin machine. PSX2 is delayed by two years
> to accommodate a total redesign.
>
Wow. How could the PSX2 be delayed if Sony never announced a release
date for it? Oh, because you were predicting the PSX2 would be out six
months after the M2 (to make it obsolete)? So instead of admitting you
were wrong, you constructed this grand fantasy that the PSX2 was delayed
because of the Saturn2. Care to speculate when the Saturn2 will be out
or what its performance will be? I mean, you've been shown wrong so
many times before, you might as well not stop now.
> M2 could have been something if it was released back in 1995; now,
> it is an outdated technology.
>
Yeah, its just more powerful than any other console on the market,
that's all.
-Todd Allendorf
tod...@alumni.caltech.edu
Leo L. Schwab
In article <19970218170...@ladder01.news.aol.com>,
BEwert <bew...@aol.com> wrote:
>"The real strength of 3DO lies in multimedia architecture/design talent" a
>Hugh Martin quote.
...Most of whom are gone.
Owing to a combination of artless, haphazard layoffs and people
resigning in disgust, nearly all of the magnificent people 3DO had for
advanced systems design are no longer there. A few people stayed behind to
stick it out with AGT (Advanced Graphics Technology, a/k/a Adrian Greg
Toby), and though a handful of these people are absolute top notch, I
personally don't feel there are enough of them to do the scale of work
necessary to bring the 128-bit MX system to a proper fruition. There are
certainly not enough software people remaining...
>"As far as 3DO's 128-bit system development is concerned, Toby Farrand
>said, "We're very far along."
>
I wouldn't agree here. Based on what I know of MX's design
details, if I were planning on buying the technology, there are a *lot* of
design changes I would *insist* on...
_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_
Leo L. Schwab -- Digital Spellweaver ew...@best.com ..or..
\_ -_ http://www.best.com/~ewhac/ ew...@well.com
O----^o Recumbent Bikes: The Only Way To Fly. (pronounced "EH-wack")
"Work FOR? I don't work FOR anybody! I'm just having fun." -- The Doctor
BEwert
Noone's Home spewed:
>>
[SNIP]
> in addition to a separate 3D setup engine to free
>the processing power of the two PowerPC's.
Wow, what a messy architecture. Coder's worst nightmare. No wonder
why developers are calling M2 the most difficult console to develop for,
ever.
[MORE SNIPPAGE]
M2 is too little, too late. It is not going to be able to match
upcoming Lockheed's machine for Sega(this Lockheed machine supposely
exceeds Model3's performence). I think Sony canned original PSX2 design
because of this Lockheed Martin machine. PSX2 is delayed by two years
to accommodate a total redesign.
<<
So the Saturn's architecture is great, but M2's isn't?
Everything else is your own made up BS, which is par for the course.
I was trying to bring something real info in from a mag most people here
probably don't read very often. Nothing more, nothing less. You can have
your Saturn, your PSX, your PC whatever. We don't really care.
Quote your sources, use your real name, or go home to mommy.
Nicholas Albright
In article <5eg42o$8...@lana.zippo.com>, <Nobody's_Perfect> wrote:
>
>> Its the easiest console
>>to develop for. The reasons are 1) A unified memory architecture. No
>>more shuffling data from memory pool to memory pool crap (like on the
>>PSX or Saturn) which makes resource management a nightmare and kills
>>performance.
>
<Laugh!> Oh My Lord! Oh Wow, you are so funny. Thanks for the good
laugh. I Love the way you back things up. It's great. I know I
shouldn't do this, as it breaks all rules you have, but I'm going to back
up my belief that UMA = good, not bad. (Performance wise) (This is
easier for me than you, as I actually have sources to draw from ;)
First of all, if UMA is what you say it is, why is SGI using it in it's
newest line of workstations? (Yeah, we all know SGI cuts all those
corners ;) Next, let me just back this up with a little bit from SGI.
-----Begin UMA Quote-----
High-Performance Unified Memory
Unlike typical PC and workstation architectures, where data must be
transferred between graphics, video, and imaging memory located on separate
boards, all O2 data resides in main memory where every engine has direct
access to it. With UMA, system memory, frame buffer, z-buffer, texture
memory, rendering memory, image memory, and video memory are all the
same. No longer do dedicated pools of memory exist throughout the
computer to be utilized only at specific times; with O2, memory is
automatically reallocated according to need. Expanding texture memory
becomes a matter of simply buying a standard memory upgrade.
-----End UMA Quote-----
Thanks for the laughs nobody. ;)
-Nick
--
P.S. Any spelling mistakes are not those of the author, as he can't spell. :)
j h woodyatt
Nobody's_Perfect wrote:
> Todd Allendorf wrote:
>[...]
> >
> >YOU don't see them because $10,000 MPEG encoding systems aren't sold at
> >the Kaybee Toy Stores that you frequent.
>
>[...]
That's because yer a mo-ron. The product's name was MPEGExpress, and I
personally saw it attracting crowds at MacWorld Expo last year. It's been
advertised as a 3DO product for at least a year on the company's website.
>[...]
> >Every developer I've talked to has said the opposite.
>
>Aren't you supposed to be a college student?
>[...]
Heh heh. Heh. The word 'alumni' appears to be throwing you for a loop,
isn't it son? Here, have a prune.
--
j h woodyatt <j...@wetware.com>
>>>>> Insert Epigram Here <<<<<
Leo L. Schwab
Remember when I warned you about taunting Todd Allendorf? I hope
your asbestos shorts are up to snuff...
In article <5eg42o$8...@lana.zippo.com>, <Nobody's_Perfect> wrote:
>>1) You don't know how much the BDA costs.
>
> It better cost less than $50.
>
>> 2) The "Phillips chip" doesn't do MPEG or audio.
>
> So what? Those ASICs are available cheaply.
>
I think I've got it. The two hemispheres of his brain are
operating independently, without knowledge of each other.
First he says BDA had better cost less than $50. *Then* he says
that extra ASICs are available cheaply to make up for the Philips chip
lacks.
Let's say BDA costs $70 (I really have no idea). For that, you
get a rendering engine, DSP, and an MPEG decoder. Let's say that you can
get the Philips chip for $50. Now: How much *more* would have you to
spend to add DSP and MPEG capabilities to a system using the Philips chip?
More than $20?
BDA is an integrated solution. It's just plain cheaper.
>>If you even knew what a setup engine was, you would know it is
>>transparent to programmers.
>
> Yes, it is transparent with proper math library. But if you want to
> squeeze out a console's maximum performence, then you have to bypass
> the standard math library issued by the console maufacturer and write
> your own.
Uh, no...
The setup engine is a chunk of hardware inside BDA. It is wired
directly to the triangle engine. It is the *bare metal* as far as
software is concerned.
>>Every developer I've talked to has said the opposite.
>
> Aren't you supposed to be a college student?
>
*snicker!*
Look up the word 'alumni', or its singular, 'alumnus'. Here, let me
help you:
http://gs213.sp.cs.cmu.edu/prog/webster?alumnus
Since graduating from CalTech (you *have* heard of CalTech,
haven't you?), Todd worked at 3DO, and spent most of his time writing tools
to make M2 dance. Todd also did some direct developer support. Todd *IS*
the proverbial Horse's Mouth.
_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_
Leo L. Schwab -- Digital Spellweaver ew...@best.com ..or..
\_ -_ http://www.best.com/~ewhac/ ew...@well.com
O----^o Recumbent Bikes: The Only Way To Fly. (pronounced "EH-wack")
UniCode: Your assurance that every other byte is wasted.
Leo L. Schwab
In article <5eh18v$4...@shellx.best.com>, Leo L. Schwab <ew...@best.com> wrote:
>In article <19970218170...@ladder01.news.aol.com>,
>BEwert <bew...@aol.com> wrote:
>>"As far as 3DO's 128-bit system development is concerned, Toby Farrand
>>said, "We're very far along."
>>
> I wouldn't agree here. Based on what I know of MX's design
>details, if I were planning on buying the technology, there are a *lot* of
>design changes I would *insist* on...
>
No sooner do I post this than I discover through my Clandestine
Information Network (pat. pend.) that MX's most glaring flaw has been
fixed. MX's chances of finding a customer just went up a *heck* of a lot...
Most of my other problems with MX are teensy weensy little details
that nobody but me seems to care about, which means they're probably
unimportant.
They still have a lot of other non-technical problems to surmount,
however. For the sake of my friends still there, I hope they can meet the
challenge.
_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_
Leo L. Schwab -- Digital Spellweaver ew...@best.com ..or..
\_ -_ http://www.best.com/~ewhac/ ew...@well.com
O----^o Recumbent Bikes: The Only Way To Fly. (pronounced "EH-wack")
"You don't see a zucchini cartel, do you?"
BEwert
>"The real strength of 3DO lies in multimedia architecture/design talent"
a
>Hugh Martin quote.
...Most of whom are gone.
Owing to a combination of artless, haphazard layoffs and people
resigning in disgust, nearly all of the magnificent people 3DO had for
advanced systems design are no longer there.
<<
That's hy I put the "Hugh Martin quote" part in ; )
It's amazing how a good thing can get screwed up. Hopefully Matsushita
won't ef up the M2 too bad.
E
Nigel Parker
On 20 Feb 1997, it was written:
: In article <5ehrqj$2...@bobcat.ent.ohiou.edu>, albr...@bobcat.ent.ohiou.edu says...
:
: >> No, UMA hurts overall performence. UMA was chosen to save money.
: >
: > <Laugh!> Oh My Lord! Oh Wow, you are so funny. Thanks for the good
: >laugh. I Love the way you back things up. It's great. I know I
: >shouldn't do this, as it breaks all rules you have, but I'm going to back
: >up my belief that UMA = good, not bad. (Performance wise) (This is
: >easier for me than you, as I actually have sources to draw from ;)
: >
: > First of all, if UMA is what you say it is, why is SGI using it in it's
: >newest line of workstations? (Yeah, we all know SGI cuts all those
: >corners ;) Next, let me just back this up with a little bit from SGI.
:
: SGI's UMA implementation has a bandwidth of 1.6GB and couple MB of cahce,
: enough not to hurt overall performemce. M2's UMA implementation has a
: bandwidth of only 500MB and no L2 cache, not enough to get a decent
: perforemce out. That is the differemce betweem SGI and M2. M2's UMA is
: flawed.
The PPC602 has reasonably large L1 caches (IIRC). M2 has a fast memory
bus and fast memory (making the need for L2 cache very much less
important).
M2 strikes me as being very well designed. The only dodgy factor was the
addition of the second PPC602. But I seem to remember someone from 3do
saying (a few years ago) that M2 was always designed to be scaleable
(whether or not they meant in this fashion - I don't know). In any case a
second CPU can be very useful - it opens up multithreading options or, at
the other extreme, if the OS is well written then it would behave as if
there were only a single CPU.
After all these years M2 still gives the impression of cutting edge
technology, still full of promise (IMHO).
Wolfson Court
Clarkson Road
CAMBRIDGE
CB3 0EH
Tel: 01223 338892 (messages only)
Johan Eriksson
TA> Yeah, its just more powerful than any other console on the
TA> market, that's all.
It's not. 'Cause it's not on the market (yet). ;-D
/Johan
Nigel Parker
On 21 Feb 1997, it was written:
: Nigel says...
:
: >: SGI's UMA implementation has a bandwidth of 1.6GB and couple MB of cahce,
: >: enough not to hurt overall performemce. M2's UMA implementation has a
: >: bandwidth of only 500MB and no L2 cache, not enough to get a decent
: >: perforemce out. That is the differemce betweem SGI and M2. M2's UMA is
: >: flawed.
: >
: >The PPC602 has reasonably large L1 caches (IIRC).
:
: How large is L1 cache in PPC602? 8KB? 16KB?
Is that supposed to suggest that it isn't big enough?
I would guess that it has 16k data and 16k instruction. Seems to be the
norm nowadays.
: > M2 has a fast memory
: >bus and fast memory (making the need for L2 cache very much less
: >important).
:
: But not fast enough. Note that some video memory types have a transfer
: rate of 800MB/s, and the video operation does not eat up system RAM's
: bandwidth in a system with separate system RAM and VRAM.
:
: On a typical PC system that uses 64bit data path for system memory and
: 128bit VRAM or WRAM video card, there is more than 1GB/s of memory bandwidth
: available.
Have you swallowed a computing dictionary, or what?
What are you talking about? I was saying something about L2 cache being
less important for a fast memory architecture and you've started spouting
about PC video cards?
If the M2 video system works at 16bit colour and at TV resolution then
what's the problem? Why have 128bit VRAM (?) and 1GB/s bandwidth when you
don't need it?
You don't seem to understand that M2 is designed to do nice 3d graphics,
but to be cheap. It seems reasonable, therefore, that the memory system
is optimised to push around a screenful of data per second and not a
single bit more!
Comprendez?
JAMES MARTIN
Distribution:
Nobody's_Perfect wrote:
: But M2 does not work as it was intended.
The only intention M2 has is to create games that are fun play and nice to
look at.
Only time will tell if this goal is achievable.
Of course, you're too much of a pig-head to admit when you're wrong.
Todd Allendorf
Nobody's_Perfect wrote:
>
> >Is that supposed to suggest that it isn't big enough?
>
>
> >I would guess that it has 16k data and 16k instruction. Seems to be the
> >norm nowadays.
>
> PPC602 will suffer from heavy cache miss ratio, which forces it to go get the
> data from SystemRAM, which has to split its memory bandwidth between half
> a dozen devices(remeber UMA discussion we had before?).
It is true that a larger cache is always desirable, but it is not hard
for game programmers to design small inner loops (they usually do
anyway) for critical portions of their code.
M2 ships with a graphics library (with full source) that makes extremely
efficient use of the caches.
As for the UMA discussion, you lost that one. A UMA is almost always
more efficient for all but the most contrived cases. On a Saturn or
PSX, this is a typical scenario in a Doom-style game.
You need to get some extra data (sound or graphics) off the disc. You
must tell the I/O chip to load the data into the CD-ROM cache. Then
you must tell the memory controller to move the data to whatever memory
pool its needed in, let's say audio memory. But wait, audio memory is
full (even though there is still free RAM elsewhere), so you must first
free up something else. You have to check with the audio processors
that something can be moved (you don't want to move something that's
being played at this moment). You've found a piece of memory that can
be used, so you can either delete it or move it somewhere else. If you
delete, you may need it again right away (maybe the person will walk
back into the room that the data was needed for) so maybe you should
move it to main memory. If you can find a space in main memory, then
you copy the data there. Once the copy is complete, then FINALLY, you
can copy your new data from the CD-ROM cache to the audio memory. Once
copying is done (since neither the Saturn or PSX have multi-tasking, the
programmer must continually poll the memory unit for completion of the
task or wait and waste valuable processing cycles), the audio processor
then can be allowed to use the audio data. This little exericise has
moved three times the amount of data and had to talk to four different
subsystems to do it. All these copies and syncronizations must be
directly handled by the programmer.
On a UMA system like the N64 or M2, the data is copied from the external
device to any available memory. Once the copying is done (or in the
case of the multi-tasking M2, once enough of the data has been copied)
the processor can directly use the data where it is. That's it. The
data is copied only once and only the two sub-systems that actually need
the data have to be involved.
This is why a system like the Saturn is so much more difficult than a
system like M2, and a UMA in all but the most contrived cases is much
more efficient.
>
> I am pretty sure all the FXs M2 proclaim to have eat up tons of bandwidth;
> you can never have enough memory bandwidth.
>
By your arguments, the N64 should really be hurting (since it has
approximately the same memory bandwidth as the M2). But wait, it was
designed by SGI, those guys you claim know how to make an effective
UMA. So according to you, SGI's engineers are both idiots and
geniuses. A contradiction in Nobody's arguments? How can this be?
> >You don't seem to understand that M2 is designed to do nice 3d graphics,
> >but to be cheap.
>
> But M2 does not work as it was intended.
>
Okay, everybody who has programmed (or even seen) an M2, raise your
hand. What? You've never even seen an M2? Wow, that means you don't
know what the fuck you are talking about. Imagine that!
-Todd Allendorf
tod...@alumni.caltech.edu
CJK
In article <5epopp$n...@main.freenet.hamilton.on.ca>,
ab...@freenet.hamilton.on.ca (JAMES MARTIN) wrote:
= Distribution:
=
= Nobody's_Perfect wrote:
=
= : But M2 does not work as it was intended.
=
=
= The only intention M2 has is to create games that are fun play and nice to
= look at.
=
= Only time will tell if this goal is achievable.
=
= Of course, you're too much of a pig-head to admit when you're wrong.
Hey now. Let's not be insulting. I know _plenty_ of pigs that have no
problem admitting that they are wrong.
"When in doubt, mumble."
* c...@why.net *
http://www.why.net/users/cjk