🚀 go-pugleaf

j-h...@ux4.cso.uiuc.edu (James Hogan) writes:
>gme...@lindblat.cc.monash.edu.au (GM Enierga) writes:
>>findings of the demo on the net, and I believe you even replied to that post too
>>- agreeing with some of his negative comments.

>Which findings were these?  I've responded to dozens of messages over the 
>past week or so

	Well, it was the guy (sorry I don't have his name) who did a pretty
thorough report on how the demo ran: its cinematics, frame rates, access times
from the CDROM, educated guesses, and many many more things... like I said,
pretty thorough.  And you did say a few things in direct answer to his comments
- I remember cause I commented on yours. ;-)


> (as have you, for that matter)  :)

	*grin* Funnily enough I have been "stuck" on this newsgroup for quite a
while ever since I offered my opinions on this topic.
	But I can't say it hasn't been fun, informative or interesting. :-) 


>>>Did it really?  It's been a while.  :)
>>>I believe he was the producer, while the designer was Joel Manners (?).  
>>>At any rate, I got the impression that he wasn't as invovled in the 
>>>design process as he was in the rest of the WC games.  (I could be wrong.)

>>	Well, since you are the official FAQ writer (thanks by the way) you
>>should know more about this than I do.

>:)  I really haven't done a whole lot on Priv for the WC3 FAQ, though.  

	Well, why would you?  It is a WC3 faq, not an OSI faq.


>Plenty of research on WC3 and a decent amount on Armada (since it was not 
>yet released when the first version of the FAQ was), but not much on the 
>older games.  (I'm not the same guy who wrote the Privateer FAQ, if 
>that's what you were thinking.)  

	Actually, I did not even know there was a Priv. faq... but it raises
an interesting point: does ANYONE out there know for SURE whether Chris Roberts
simply attached his name to "Privateer" and that's it or whether he had a hand
in making the game?


>>	But if what you said is true - that he wasn't as involved - then it
>>makes a whole lot of sense that the game did not turn out to be what it should
>>have been.  Chris Roberts is "just off" the level occupied by Mr. Meier... and
>>it is things like this which have kept him "just off", unfortunately.  After
>>seeing what he came up with when he did Times of lore, Space Rogue, and WC
>>series SO FAR, one knew that he was/is destined for cyberspace greatness. :-)

>I never played TOL or Space Rogue, though most people I've talked to who 
>played them liked them.  And then there's Bad Blood... :)  not _horrible_, 
>but it never really held my interest for a long time.

	I guess it just depends what games one is into.


>>>I wish they would have recognized the AI as a problem and fixed it... 
>>>then it really could have been a much better game.

>>	Well the strategy could have been improved too.  Simply doing basic
>>planet-hopping, making better ships and looking for the enemy carrier is not
>>THAT big of a deal.

>Agreed... by "AI" I meant the combat as well as strategic intelligence.
>  |      James P. Hogan      | Check out the REVISED WC3 FAQ, version 1.10! |

	Ditto.

							- Falcon

In article <3bdqva$7...@vixen.cso.uiuc.edu>,
James Hogan <j-h...@ux4.cso.uiuc.edu> wrote:
>>I believe there is no difference between VLB and PCI, performance-wise, on
>>a 486.  If you have the choice, however, go for the PCI.
>
>Why?  :)  (Not trying to be a smart-ass, I'm just not extremely 
>knowledgable about video cards.)  Any reasons other than the fact that 
>you'll be able to keep the PCI if you upgrade to a Pentium?

	Actually, PCI has some performance edge over VLB.  PCI can
configure IRQs, DMAs and addresses for your PCI cards if there is a
conflict with ISA cards already plugged-in.  A PCI IDE controller can
also sense what hard disks are connected (but these have to be recent
IDE drives).  PCI also uses burst DMA so has greater bandwidth than VLB.

TRH Sends

>>I believe there is no difference between VLB and PCI, performance-wise, on
>>a 486.  If you have the choice, however, go for the PCI.
>
>Why?  :)  (Not trying to be a smart-ass, I'm just not extremely 
>knowledgable about video cards.)  Any reasons other than the fact that 
>you'll be able to keep the PCI if you upgrade to a Pentium?

I have a PCI 486,a and I chose to pay $100 extra for the PCI config
Being able to keep the PCI periferals when you upgrade to Pentium is a nice
option. When I get the money together, I'll swap out the motherboard, plug
in my PCI video card and hard drive controller, and sell (hopefully) the old
486 motherboard.

Isn't PCI supposed to be able to run at 60 or 66 MHz?  Or was this only for
PCI 2.0.  It makes me wonder if the big win for PCI/486 might be with the 
DX 50, as the bus could then fly at 50 Mhz.  At any rate, my 486DX2-66 PCI 
(Genoa Phantom W32/p 2meg) kicks butt over all the 486DX2-66 VLB (Western
Digital WG-1000VL) machines where I work.  This from eyeballing how well
System Shock demo works on both machines.  The slower ones at work here are
1/2 a year younger than mine at home, so hard drive speeds should be equal.
The real difference is 256kb cache at home vs 128kb at work.

(Anybody got 2 cents I can borrow?)

-Claus Schwinge

GM Enierga (gme...@lindblat.cc.monash.edu.au) wrote:

: >If you have a DX-50,

: 	Nope... and btw - is a DX-100 a dx-99 really (ie a DX4)?  In other
: words, is it based on the DX33 or the DX25?
I believe that's correct (based on dx33 I believe)

: > Local bus video card,

: 	Nope... and again, for a 486 computer, WHICH is better? VLB or PCI?

One of these tech wizzes is going to have to help us out on that one :)

: And is it likely that when the 686 will be released they will bring out a
: new chipset as they brought out the PCI for the Pentium chips?

I read a while ago that someone was working on a 128 bit card, so maybe
they'll have a new chip set with that. If one of the video card
manufacturers is reading, a chipset designed with games in mind would be
quite nice.


: > your set for SVGA graphics
: >in dogfights where a capital ship is not present. The engine in the demo is
: >unoptimized. I actually think the engine is date before Armada was released,
: >so it unoptimized, as they admit in the readme file. So who knows how much
: >faster it's going to be on final release?

: 	I think James the FAQ writer can really answer that - he's more in the
: know than most of us here I suspect.


: >I agree, my DX-50 is going be lasting me at least another 2 years,
: >regardless of what OSI or any other company does.

: 	How long have you had it?  I hope you are right though... even though
: games like Nascar are being released, at the moment it really does not worry
: me since Nascar (I hear) cannot run fully on a P90 even.  Which reminds me:
: what specs does EA's "US Navy Fighters" (or whatever) need?

What I meant is that I'm not going to upgrade until I'm ready. If every
company releases games in say one years time that all require P90's as the
minimum requirements, then the software companies will lose out on my
business until I upgrade. The games will still be on the shelves afterall,
probably for a reduced price.

: 	Well, I FINALLY had a look at the demo too.  I was very impressed -
: especially with the acting, it really did make me feel like watching the
: trailer for a top new sci-fi movie... I especially liked Tolwyn's (McDowell)
: and that other redhead-spiked-hair guy's (not Maniac, but the one who says in
: point form something about not having R&R for so long) performances...
: but I digress.  The Kilrathi looked better than I expected, but somehow those
: "flat elongated lips" still remind me of those Dark Crystall muppets...  I
: think there is still room for improvement (or is it too late?  How about
: touching up with comp. graphics like they did in the Crow?).

The final game will be in SVGA, maybe it won't be quite as noticable, but I
really am impressed with what I see. I was initially against the idea of
video footage, but I am now happy with it :)

: 	At least they remembered to keep the Emperor's bad eye. :-)
Yes

: 	If that is true, then perhaps the complaints of "people like me" have
: not been in vain.  About Bio-forge though, is it REALLY possible to make a
: game like that "totally interactive"?
In CD-rom format, I believe it is. Theyt have the space in which to make the
world big, and interactive. They can have multiple plot lines. I don't know
if Bioforge will be as interactive as they like, it was initially designed
for floppy release, but a Bioforge 2 could possibly be totally interactive.

: 	Within reason, I guess.  After running the demo on my DX33 ISA ET4000
: Tseng labs card (yes, rather old by gaming standards), I found it quite
: satisfactory... even though I could NEVER force-run the SVGA combat screens
: no matter what key I had pressed (and yes I had preloaded my Vesa driver).
: Is it some kind of ALT key combo, or will the program NOT run it cause it
: exceeds some kind of "minimum SVGA requirement" value?
Hit alt-3, it should be in SVGA then. Do you get SVGA with other games that
have SVGA capabilities?

: 	By the way, did anyone have problems with their SB type card?  I think
: my volume chip on the card went haywire after running the demo for the 1st time.
No worries here, with my SoundBlaster. Although I do have it running through
my stereo.

: 	And my MOST sincerest thanks to the person who uploaded the demo.  Quite
: readily the best demo I have ever downloaded. :-)

Bloody well right:)
-- 
.__________________________________________________________________________.
|Internet: pfle...@scu.edu.au                                             |
| One hour from now WING COMMANDER 3 will be one hour closer to release    | 
|__________________________________________________________________________|

>:       Nope... and again, for a 486 computer, WHICH is better? VLB or PCI?

For all the bench marks I've seen PCI kill vesa for video but not disk access. 
I would think that that will come though.  
 

>: And is it likely that when the 686 will be released they will bring out a
>: new chipset as they brought out the PCI for the Pentium chips?

They allready have new chip sets in the works, both PCI and VESA LB will have 
new versions soon.  Hardly any Mother boards are even shipping with the full 
64 bit PCI yet.  There is almost no hardware for it yet.  

>I read a while ago that someone was working on a 128 bit card, so maybe
>they'll have a new chip set with that. If one of the video card
>manufacturers is reading, a chipset designed with games in mind would be
>quite nice.

There is already a 128 bit card out.  It's called the Imagine by #9. It fits, 
I believe in a 32 bit Vesa or PCI slot (different versions of the card 
obviously).  It is mainly aimed at CAD apps.

>: > your set for SVGA graphics
>: >in dogfights where a capital ship is not present. The engine in the demo is
>: >unoptimized. I actually think the engine is date before Armada was released,
>: >so it unoptimized, as they admit in the readme file. So who knows how much
>: >faster it's going to be on final release?

I was very impressed on my dx50 with Isa, et4000, old video card.  It was 
smooth enough to blast bad guys.  Although with that AI I could have done at 2 
FPS with a little Force.

>: >I agree, my DX-50 is going be lasting me at least another 2 years,
>: >regardless of what OSI or any other company does.

Not Mine I hope.  P90, here I come.  Hopefully right after Xmas season.  

>:       How long have you had it?  I hope you are right though... even though
>: games like Nascar are being released, at the moment it really does not worry
>: me since Nascar (I hear) cannot run fully on a P90 even.  Which reminds me:
>: what specs does EA's "US Navy Fighters" (or whatever) need?

It's kind of funny that companies are anticipating the 686 already, in the 
sence that for the first time Intel is seriously planning to move away from 
the 8086 structure and the target market and design features of the 686 are in 
question right now.  On the otherhand I don't think it was all that hard for 
EA and Papyrus to design there games above current hardware levels.  

>>:       Well, I FINALLY had a look at the demo too.  I was very impressed -
>: especially with the acting, it really did make me feel like watching the
>: trailer for a top new sci-fi movie... I especially liked Tolwyn's (McDowell)
>: and that other redhead-spiked-hair guy's (not Maniac, but the one who says in
>: point form something about not having R&R for so long) performances...
>: but I digress.  The Kilrathi looked better than I expected, but somehow those
>: "flat elongated lips" still remind me of those Dark Crystall muppets...  I
>: think there is still room for improvement (or is it too late?  How about
>: touching up with comp. graphics like they did in the Crow?).

>The final game will be in SVGA, maybe it won't be quite as noticable, but I
>really am impressed with what I see. I was initially against the idea of
>video footage, but I am now happy with it :)

I anticipate much better graphics in the final version.  I could be wrong, but 
I doubt origin will release this with crisp graphics, since it is one of the 
major selling points and it certainly can be done.  i thought the acting was 
first rate in comparison with EVERY other game out there and many movies.  I 
loved the "make your peace" part.

>:       If that is true, then perhaps the complaints of "people like me" have
>: not been in vain.  About Bio-forge though, is it REALLY possible to make a
>: game like that "totally interactive"?
>In CD-rom format, I believe it is. Theyt have the space in which to make the
>world big, and interactive. They can have multiple plot lines. I don't know
>if Bioforge will be as interactive as they like, it was initially designed
>for floppy release, but a Bioforge 2 could possibly be totally interactive.

I think the bigest problem with interactive games is that it would take so 
much vision to guess every thing that every person might try to do in a game.  
Thats where the game begins to break down.  But they keep trying and 
occasionally the fruits of there labor are worth picking.

>:       By the way, did anyone have problems with their SB type card?  I think
>: my volume chip on the card went haywire after running the demo for the 1st
>time.
>No worries here, with my SoundBlaster. Although I do have it running through
>my stereo.

Sorry SB16 worked fine.   

>:       And my MOST sincerest thanks to the person who uploaded the demo.  Quite
>: readily the best demo I have ever downloaded. :-)

>Bloody well right:)

Certainly is.
PDeMaio

Is that a light at the end of the tunnel or just an oncoming train?

pfle...@scu.edu.au (Nightfox) writes:
>GM Enierga (gme...@lindblat.cc.monash.edu.au) wrote:
>: And is it likely that when the 686 will be released they will bring out a
>: new chipset as they brought out the PCI for the Pentium chips?

>I read a while ago that someone was working on a 128 bit card, so maybe
>they'll have a new chip set with that. If one of the video card
>manufacturers is reading, a chipset designed with games in mind would be
>quite nice.

	Is this a 128-bit graphic card or something else?


>: >I agree, my DX-50 is going be lasting me at least another 2 years,
>: >regardless of what OSI or any other company does.
>: 	How long have you had it?  I hope you are right though... even though
>: games like Nascar are being released, at the moment it really does not worry
>: me since Nascar (I hear) cannot run fully on a P90 even.  Which reminds me:
>: what specs does EA's "US Navy Fighters" (or whatever) need?

>What I meant is that I'm not going to upgrade until I'm ready. If every
>company releases games in say one years time that all require P90's as the
>minimum requirements, then the software companies will lose out on my
>business until I upgrade. The games will still be on the shelves afterall,
>probably for a reduced price.

	:-) Fair enough... I am of a similar mind, but I keep forgetting
something which I know not the answer at this time since nothing has caused it:
it is called "Temptation". :-)


>: Re: WC3 demo...
>The final game will be in SVGA, maybe it won't be quite as noticable, but I
>really am impressed with what I see. I was initially against the idea of
>video footage, but I am now happy with it :)

	I was really wondering how "interactive" video footage can be; after
seeing it, it feels it will be good, but again I will have to see it for
myself before judging it.


>: 	If that is true, then perhaps the complaints of "people like me" have
>: not been in vain.  About Bio-forge though, is it REALLY possible to make a
>: game like that "totally interactive"?

>In CD-rom format, I believe it is. Theyt have the space in which to make the
>world big, and interactive. They can have multiple plot lines. I don't know
>if Bioforge will be as interactive as they like, it was initially designed
>for floppy release, but a Bioforge 2 could possibly be totally interactive.

	Something that needs clarification here: does the term "interactive"
mean one can step outside the plotline of a game like in Mechwarrior, Space
Rogue or Privateer (though these are poor examples, since you only ever
basically traded and fought and nothing else - no character development away
from the main game story... Arena (Elder Scrolls V1) I think was SUPPOSED to
do this but I suspect they did not - I am not sure)?  To me, that is what total
interactivity means.
	So basically I need to know if my definition is the same as Origin's.


>: 	Within reason, I guess.  After running the demo on my DX33 ISA ET4000
>: Tseng labs card (yes, rather old by gaming standards), I found it quite
>: satisfactory... even though I could NEVER force-run the SVGA combat screens
>: no matter what key I had pressed (and yes I had preloaded my Vesa driver).
>: Is it some kind of ALT key combo, or will the program NOT run it cause it
>: exceeds some kind of "minimum SVGA requirement" value?

>Hit alt-3, it should be in SVGA then. Do you get SVGA with other games that
>have SVGA capabilities?

	I tried ALT-3.. or was it ALT-F3?  I must go back and try again.  Does
anyone have some kind of list with what all the keys not in the demo's text file
do?  James... how about including something like this in your next FAQ? :-)
	Yes, my card does have SVGA capabilities... I play Links 386 with the
SVGA courses.


>: 	By the way, did anyone have problems with their SB type card?  I think
>: my volume chip on the card went haywire after running the demo for the 1st time.
>No worries here, with my SoundBlaster. Although I do have it running through
>my stereo.

	I should have been more precise: an SBAWE32 to be exact.

					- Falcon

In article <3bdu68$p...@tuba.cit.cornell.edu>, ak...@crux5.cit.cornell.edu (Allen Kim) writes...
>j-h...@ux4.cso.uiuc.edu (James Hogan) writes:
>>ak...@crux5.cit.cornell.edu (Allen Kim) writes:
>>>I believe there is no difference between VLB and PCI, performance-wise, on
>>>a 486.  If you have the choice, however, go for the PCI.
>>Why?  :)  (Not trying to be a smart-ass, I'm just not extremely 
>>knowledgable about video cards.)  Any reasons other than the fact that 
>>you'll be able to keep the PCI if you upgrade to a Pentium?
>Actually, that's it.
To tell the truty, you won't even be able to keep the motherboard if you
want to upgrade from a 486 to a Pentium.  If you upgrade to the P60, you will
need an extra fan and a a heat sink, because the chip runs so hot.  If you
upgrade to the P90 or P100, you will need a new board that supplies 3.3V
to the chip rather than the 5.5V required for the 486's and P60's (DX4's
however, run on 3.3V as well)

The best choice of bus should follow your choice of graphics card.  Most
cards come in either VLBus or PCI variety, and the speed results in each
format are published in places like Computer Shopper.  It depends on the
manufactuer of the card on which format is faster, and they are usually
pretty close to the same speed.

K

(P.S., I would adivse against getting anything from Diamond, they are 
extremely fast, but have palatte problems in lower resolutions (the
ones we use for our games, unfortunately)  It is unpredictable when
you'll have a problem with a diamond card, I've even seen some of them
have problems with Windows.  Also, at my company every diamond card ordered
had to be returned for a replacement twice right after purchase until
some working ones began to come in.  Their quality control is apparently
not very good.)

In article <D025o...@scu.edu.au>, Nightfox <pfle...@scu.edu.au> wrote:

>: 	Nope... and btw - is a DX-100 a dx-99 really (ie a DX4)?  In other
>: words, is it based on the DX33 or the DX25?
>I believe that's correct (based on dx33 I believe)

	Nope...  A DX4-100 actually runs a tad faster than 100mHz
internally.  Exact, down-to-the-hertz, 3-to-1 synchronization between the
100mHz processor and 33mHz external motherboard bus is not necessary
because the DX4-100 has a 16K cache where the outgoing code can wait
until the next available motherboard 33mHz clock cycle.  So the 100mHz
CPU and motherboard are in fact running a bit asynchronously.

	That's because the CPU and motherboard both have their own clocks.
If the CPU and motherboard shared one clock then synchronization would be
important, as was in the days of the 286 and 8088.
 
TRH Sends

j-h...@ux4.cso.uiuc.edu (James Hogan) writes:
>Already in the works.  :)  Try Alt-4 if nothing else works.  After that 
>try Alt-3... that should speed it up on some video cards (particularly if 
>you have a good VLB one), though it may slow it down on others.  
>  |      James P. Hogan      | Check out the REVISED WC3 FAQ, version 1.10! |

	What EXACTLY does the ALT-4 combination supposedly do?  And the alt-1
& 2 combos?
	So what version are you going to call the new one? :-)

						- Falcon

In article <3bm8d9$p...@harbinger.cc.monash.edu.au>,
GM Enierga <gme...@lindblat.cc.monash.edu.au> wrote:
>j-h...@ux4.cso.uiuc.edu (James Hogan) writes:
>>Already in the works.  :)  Try Alt-4 if nothing else works.  After that 
>>try Alt-3... that should speed it up on some video cards (particularly if 
>>you have a good VLB one), though it may slow it down on others.  
>
>	What EXACTLY does the ALT-4 combination supposedly do?  And the alt-1
>& 2 combos?
>	So what version are you going to call the new one? :-)
>
>						- Falcon
>
Origin says ALT-3 does single-buffering, and ALT-4 does triple-buffering.
ALT-1 is equivalent to ALT-3, but in VGA, and ALT-2 is equivalent to ALT-4.
It's possible that ALT-1 uses double-buffering, but that's similar enough
to single-buffering that it doesn't really matter for this discussion.

Anyway, here's what it boils down to:  ALT-3 (and ALT-1) produce the frame
in memory, then stuffs it to the video card as fast as the bus can handle
it.  This works great if you have a fast video card and local bus, and
requires minimal CPU time.  It's called single buffering because only a
single frame is stored in memory at a time.

Triple-buffering is much more complex.  Suppose the computer has just finished
drawing a frame.  The CPU calculates the next frame, and then compares it to
the frame it just drew.  Then it ONLY draws the pixels that have changed to
the monitor.  This is impressive stuff, and is extremely efficient if you
have a slow card or bus, but a fast computer.  However, it does require
significantly more CPU time.  Origin is the only company I know of that
bothers to implement something this difficult, and my hat goes off to them
for it.  It's called triple-buffering because the previous frame, the current
frame, and the pixels to change (three items) are all stored in memory at 
the same time.

To prove how impressive a technique this is, I'll use my system as an
example.  I recently purchased a Pentium motherboard, but haven't had the
chance to put a new video card in it.  WC3 rates it:

0 for CPU
17 (!!) for my ISA Trident 8900C.

In VGA, I get 24 frames per second no matter what.
With ALT-3, I get 5 frames per second.  Always.  My video card simply cannot
accept faster rates than that.  However...
With ALT-4, I get 24 frames per second in clear sky.
                  20-24 frames per second when dogfighting, and
         at worst 7 frames per second when those Rookies run into me,
which only lasts for a frame or two anyway

All you Origin bashers: THAT is good programming.  Yet it's the kind of thing
that you could never know about just by playing the game, unless you had
a pretty good idea of how these things worked.  And they've been using this
same technique AT LEAST since the floppy version of Strike Commander!
Their only mistake was not anticipating that local bus video would make this
technique obsolete... and that problem was fixed in the CD release.

Uh, anyway, that's the difference between the video modes in WC3.  Sorry
you asked?  ;)

-Jim Little  (ji...@teleport.com)

: 	Is this a 128-bit graphic card or something else?
I got the impreesion that it was a graphic card. I really can't remember
where I read it though. Maybe they'll get getting bigger busses with every
new motherboard, until the get one that stretches from here to the moon :)

: 	:-) Fair enough... I am of a similar mind, but I keep forgetting
: something which I know not the answer at this time since nothing has caused it:
: it is called "Temptation". :-)

Aaah yes, my master plan to get around this is to purchase a game once every
3 months. WC3 December, System Shock CD Februrary/March, Bioforge June. With
occasional older game picked up on the cheap somewhere, like Lands of Lore
CD is one that I'm looking for. I think Wing Commander 3 will be good for
quite a number of replays (I'm still not bored with the demo) and before I
know it, I'll be right when upgrade time comes :)


: 	Something that needs clarification here: does the term "interactive"
: mean one can step outside the plotline of a game like in Mechwarrior, Space
: Rogue or Privateer (though these are poor examples, since you only ever
: basically traded and fought and nothing else - no character development away
: from the main game story... Arena (Elder Scrolls V1) I think was SUPPOSED to
: do this but I suspect they did not - I am not sure)?  To me, that is what total
: interactivity means.
: 	So basically I need to know if my definition is the same as Origin's.

For me an interactive movie means that I have some control over how the
story line unfolds. In Bioforge for example, some of the decisions end up
how the character turns out to be. One good thing about Pacific Strike is
how they did that with the story line as well. Repeated failures and the
moral drops, and I didn't play long enough to win.


: 	I should have been more precise: an SBAWE32 to be exact.

That I know nothing about :)
-- 
.__________________________________________________________________________.
|Internet: pfle...@scu.edu.au                                             |
| One hour from now WING COMMANDER 3 will be one hour closer to release    | 
|__________________________________________________________________________|

In article <D07Lz...@scu.edu.au>, Nightfox <pfle...@scu.edu.au> wrote:
>: 	Is this a 128-bit graphic card or something else?
>I got the impreesion that it was a graphic card. I really can't remember
>where I read it though. Maybe they'll get getting bigger busses with every
>new motherboard, until the get one that stretches from here to the moon :)

	When they say a graphics card is 24-bit, 32-bit, 64-bit, or 128-bit,
it has nothing to do with bus width.  All the above cards use a 32-bit-
datapath bus (PCI, VLB, or EISA), or 16-bit if one still uses an antiquated
ISA machine.  24-bit simply means the ability to display 16.7 million colors
at 640x480, 32-bit means 800x600x16.7m, 64-bit means 1024x786x16.7m, and
128-bit means 1248x1024x16.7m.  The only 128-bit graphics card out there
right now is the Imagine128 by Number Nine, which is bundled with Dell
Pentium-90s.  It uses the 32-bit PCI bus.

TRH Sends

th...@aloha.cc.columbia.edu (Tom  Hong) writes:
>In article <D07Lz...@scu.edu.au>, Nightfox <pfle...@scu.edu.au> wrote:
>>: 	Is this a 128-bit graphic card or something else?

>>I got the impreesion that it was a graphic card. I really can't remember
>>where I read it though. Maybe they'll get getting bigger busses with every
>>new motherboard, until the get one that stretches from here to the moon :)

>	When they say a graphics card is 24-bit, 32-bit, 64-bit, or 128-bit,
>it has nothing to do with bus width.  All the above cards use a 32-bit-
>datapath bus (PCI, VLB, or EISA), or 16-bit if one still uses an antiquated
>ISA machine.

	Me! :-)


>24-bit simply means the ability to display 16.7 million colors
>at 640x480, 32-bit means 800x600x16.7m, 64-bit means 1024x786x16.7m, and
>128-bit means 1248x1024x16.7m.

	Well, that makes everything clear... thanks for the info.  Just a final
point, the Mb of RAM on a card has nothing to do with the X-bits of a graphic
card?


>The only 128-bit graphics card out there
>right now is the Imagine128 by Number Nine, which is bundled with Dell
>Pentium-90s.  It uses the 32-bit PCI bus.
>TRH Sends

	So a Tseng LAbs ET4000-w32p is what "bits" graphic card, please?

						- Falcon

In article <3buuas$i...@harbinger.cc.monash.edu.au>,
GM Enierga <gme...@lindblat.cc.monash.edu.au> wrote:

>point, the Mb of RAM on a card has nothing to do with the X-bits of a graphic
>card?

	Hmm...  I think it does.  For example, my Diamond Stealth 64 PCI
has 2MB of Video RAM.  That means it can display 16.7 million color
palettes (2 to the 24th power = 16,777,216 colors), or 24-bits per pixel,
at 800x600 resolutions, or a total of 480,000 pixels on-screen.  So 24-bits
per pixel times 480,000 pixels is 11,520,000 bits, divided by 8 yields
1,440,000 bytes of video RAM required. But the Diamond Stealth 64's VRAM
can be expanded to 4MB (I have a feeling it would cost me an arm and a leg
to do this so I don't think I will anytime soon...  grin) to display 24-bit
color at 1280x1024 resolutions, for a total of 1,310,720 pixels on screen,
which is over twice the number of pixels on-screen than the 800x600
modes.  This means 1,310,720 x 24 bits per pixel divided by 8 to yield
3,932,160 bytes, which is more than what my present amount of Video RAM
can handle.

>	So a Tseng LAbs ET4000-w32p is what "bits" graphic card, please?

	I think the ET4000-w32p cards have a 32-bit chip on-board, yes?
I have seen them with 2MB of RAM on-board, so that means a maximum palette
(16.7 million true-color mode) at 800x600 resolutions.

TRH Sends

In article <3buuas$i...@harbinger.cc.monash.edu.au>
gme...@lindblat.cc.monash.edu.au (GM Enierga) writes:

> >24-bit simply means the ability to display 16.7 million colors
> >at 640x480, 32-bit means 800x600x16.7m, 64-bit means 1024x786x16.7m, and
> >128-bit means 1248x1024x16.7m.


Ummmm, I think 24 bit refers to the bit-plane/color depth, and that 32,
64, and 128 refer to the graphics processor bus-width(how much data it
can fiddle with at a time, internally).

The cards that support 16.7M only need 24 bit plates at any resolution.
 At 1024x768 they do NOT need a 64 bit anything, just enough memory to
store that much info, similar with 800x600x16.7M.

There are some cards with 32 bit planes--8 r,g,b, plus 8 alpha
channals.  (some esoteric value useful in something or other ;) )


Russ.

schultz...@semail.jsc.nasa.gov (Schultz, Russell) writes:

>In article <3buuas$i...@harbinger.cc.monash.edu.au>
>gme...@lindblat.cc.monash.edu.au (GM Enierga) writes:

>> >24-bit simply means the ability to display 16.7 million colors
>> >at 640x480, 32-bit means 800x600x16.7m, 64-bit means 1024x786x16.7m, and
>> >128-bit means 1248x1024x16.7m.


>Ummmm, I think 24 bit refers to the bit-plane/color depth, and that 32,
>64, and 128 refer to the graphics processor bus-width(how much data it
>can fiddle with at a time, internally).

>The cards that support 16.7M only need 24 bit plates at any resolution.
> At 1024x768 they do NOT need a 64 bit anything, just enough memory to
>store that much info, similar with 800x600x16.7M.

>There are some cards with 32 bit planes--8 r,g,b, plus 8 alpha
>channals.  (some esoteric value useful in something or other ;) )

In 32-bit or 64-bit images, the extra bits (8 or 16) are usually used
for an alpha channel or matte channel. This allows images to be
composited with a smooth edge, and to have transparency. That's what
an alpha channel gives you. It's basically a gray-scale image that
describes HOW the other bits of the image will show though.

>Russ.
--
Paul Miller            |              "What EEEEZ it man!?"
Software Engineer      | The opinions expressed by me do not necessarily
Elastic Reality, Inc.  |          reflect those of my employers.

To be clear, the number of bits that a card advertises is the internal
maximum wordlength that the video card's graphics processor can handle.  For
example, a 64 bit card can process an instruction word that is 64 bits long,
or 8 bytes, at a time.

The number of bitplanes that the card supports is a measure of how many colors
it can display.  A 24 bit color display, regardless or resolution, is able
to show 16.7MM colors at once, also called Truecolor.  32 bitplane display 
shows the 24 bit color, plus 8 alpha channels.

The two concepts are only indirectly related.  There have been 24 bit cards
on the market for ages (until recently they have also been absurdly expensive),
and I would doubt that the earliest Targa boards were more than 32 bit 
processors.

Similarly, if you got a 2MB 128bit card, you would still only get 16, rather 
than 24, bit color at 1024x768, regardless of the 128 bit processor on the
card.

If you care about the number of colors, the onboard memory is the number you
need to focus on.

Regards,