home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.sys.ibm.pc.games
- Path: sparky!uunet!cpqhou!scotts
- From: scotts@cpqhou.compaq.com (Scott Shaffer @ Compaq Computer Corp.)
- Subject: Re: Is it good to have a new kind of video card?
- Organization: Compaq Computer Corp
- Date: Mon, 16 Nov 92 21:13:00 GMT
- Message-ID: <1992Nov16.211300.3986@cpqhou.compaq.com>
- References: <1992Nov16.032617.27545@jupiter.sun.csd.unb.ca>
- Lines: 97
-
- in article <1992Nov16.032617.27545@jupiter.sun.csd.unb.ca>, y70q@jupiter.sun.csd.unb.ca (LAM) says:
- >
- > Seeing that there are a lot of audio card out here for games, I keep
- > wondered: is it possible to have a video card for games? There are
- > video cards out here specialized for Windows. Some of these card even
- > slow down video speed of DOS to favour Windows. So, maybe someone can
-
- Well, I think this discussion is going along and nobody has really hit the
- nail on the head when it comes to video cards and games. The real reason
- we don't have faster/higher resolution boards that get game support is the
- lack of a STANDARD.
-
- IBM had the reigns in the mid 80s and was able to develop what became the
- standards for things like MDA, CGA, EGA, VGA, and XT, and AT architecture.
- However, once Compaq came out with the 80386 *prior* to IBM, a new standard
- for PC architecture came into being. Once some of the video board clone
- companies came out with SuperEGA (800x600 mostly) they created thier own
- standard. Luckily for us, the SuperEGA cards never really took off, since
- IBM's VGA came out right after that. Unfortunately, that was the last
- standard for video. Now, we have hundreds of different video cards and
- standards for accessing those cards (even IBM now has XGA). The upside for
- the consumer is that so much competition keeps the prices low (I just saw
- a ATI XL24 board for $249 for example, and it does 24-bit color at 640x480).
- The downside is that software developers don't have consitent way of developing
- for the mass market. Say you are a software developer, and you have 9 months
- to write a game. Of that, you alot 1 month to develop graphics support. Now,
- you would like to support 640x480x256 colors, but since there isn't a standard
- and all the video cards don't support this mode in the same way (or at all) you
- must develop for the lowest common denominator (like 320x200x256 or 640x480x16).
- You just can't spend the resource dollars any other way.
-
- That is the bad news. The good news is that a group of video card mfg. formed
- a standards committe to develop a standard. They called this standard VESA.
- But, as often happens when competing companies try to standardize, they have
- been late with the standard and what they have is *less* than functional.
- They just published a workable spec, and its called VESA 1.2 (I think). The
- new games from 360 (Theatre of War and V4Victor) plus Links386 use this
- standard (implemented most often as a driver, but not necessarily) to
- access 640x400x256 colors (why 640x400? Because that fits in the 512KB most
- of the early SuperVGA cards had). But there is a snag even here, you see the
- VESA people decided all they needed was a standard that defined how to put
- the cards into that mode (and where video memory is). They were very short
- sighted, because Microsoft came in and designed a video spec for Windows that
- defines how to talk to the display in more realistic terms. They saw right
- away that they were going to do a lot of line drawing, fills, and other
- basic operations. They decided they should provide support for (what was
- then high-end) CAD video boards that included graphics coprocessors. Since
- Windows has become incredibly popular (for better or worse) the graphics
- coprocessor is now almost standard equipment on video boards. When you
- hear about S3 or ET4000 video people are refering to the coprocessor on the
- video board. Why not just standardize on the Windows video driver system?
- I have no idea, it sounds good to me. If there is a technical reason why
- it would be a bad idea I'm not aware of it. The windows model has the
- side benefit of abstracting hardware-dependent issues away from the
- program. The guy you mentioned the video double buffering problem
- sees this, and its not that Windows ignores this, it just doesn't have to
- deal with it since these details are handled at the driver level. This
- puts some pressure on the card mfg. to develop a driver, but this is the
- way it SHOULD be. They make the card, they should make the driver. Sierra
- shouldn't have to support all the cards, the cards should support Sierra!
-
- There is one last reason that you (in your fictional company) would be
- hesitant to develop a game for high-res video boards. Even if a nice
- standard existed for the different modes (VESA tries), you would have to
- get hard market data that leaving out the lower end of the market would
- still net you enough to keep the company going. Developing a game to
- run on 320x200 and 640x480 is not always feasible. Just think of all
- the artwork in a game like RexNebular that would have to be DUPLICATED in
- both screen resolutions. Then think of how many disks it would take to
- ship ALL the data to the users.
-
- This is why we see so many "creative" marketing ideas in the computer store
- these days. Some companies have no problem shipping SVGA only games (like
- 360). Some ship seperate boxes for SVGA,VGA and EGA (to save on disk costs).
- But most, especially the small companies, just can't afford to alienate part
- of the potential market.
-
- Is there hope? Sure. Eventually some standard will emerge that will enable
- everybody to use high-res boards. Eventually that standard will be antiquated
- and we'll go through the same thing when everybody complains ("How come there
- aren't any 2048x2048x24bit games out there?")
-
- Hopefully, the video market will go the way of the sound-card market and find
- an acceptible standard quickly. Whether you like it or not, the SoundBlaster
- is a good base standard to go by. The board is cheap (<$100) and easy to
- program to. Sure the Roland MT-32 is a better system, but it costs >$500!
-
- Whew, I'm really going on here. Enough out of me.
- I'm just trying to shed some light on a confusing subject.
-
- Scott
-
- + Scott Shaffer @ Systems Div. @ Compaq Computer Corp. +
- + scotts@cpqhou.se.eng.compaq.com + "Flee to me remote elf!" +
- --
- + Scott Shaffer @ Systems Div. @ Compaq Computer Corp. +
- + scotts@cpqhou.se.eng.compaq.com + "Flee to me remote elf!" +
-