Return to BSD News archive
Path: euryale.cc.adfa.oz.au!newshost.anu.edu.au!harbinger.cc.monash.edu.au!news.rmit.EDU.AU!news.unimelb.EDU.AU!munnari.OZ.AU!news.ecn.uoknor.edu!solace!nntp.uio.no!news.cais.net!newsfeed.internetmci.com!in2.uu.net!news.artisoft.com!usenet
From: Terry Lambert <terry@lambert.org>
Newsgroups: comp.os.linux.development.apps,comp.os.linux.development.system,comp.os.linux.x,comp.os.linux.hardware,comp.os.linux.setup,comp.unix.bsd.386bsd.misc,comp.unix.bsd.bsdi.misc,comp.unix.bsd.netbsd.misc,comp.unix.bsd.freebsd.misc
Subject: Re: Why to not buy Matrox Millennium
Date: Thu, 18 Apr 1996 14:10:56 -0700
Organization: Me
Lines: 115
Message-ID: <3176AFE0.28146F7@lambert.org>
References: <4kfkb2$dgs@coyote.Artisoft.COM> <stephenkDpoDrJ.177@netcom.com> <3170348D.4496D9F1@lambert.org> <stephenkDq2BCK.B40@netcom.com>
NNTP-Posting-Host: hecate.artisoft.com
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
X-Mailer: Mozilla 2.01 (X11; I; Linux 1.1.76 i486)
Xref: euryale.cc.adfa.oz.au comp.os.linux.development.apps:14555 comp.os.linux.development.system:21612 comp.os.linux.x:29565 comp.os.linux.hardware:36612 comp.os.linux.setup:50934 comp.unix.bsd.386bsd.misc:647 comp.unix.bsd.bsdi.misc:3270 comp.unix.bsd.netbsd.misc:3098 comp.unix.bsd.freebsd.misc:17542
Stephen Knilans wrote:
] >People who do not run XFree86 don't care that their interfaces
] >aren't documented. They use the proprietary X severs that
] >came with their proprietary OS.
] >
] >An undocumented interface is *not* a good excuse for these people
] >as to "Why to not buy Matrox Millennium"... the excuse does not
] >bear (and should not) on these peoples purchase decisions.
]
] How would YOU like to buy a car that runs on some material that
] only that one company can give you, and that you can't get
] elsewhere? That SHOULD bear on any purchase decision.
The situation is not analogous. XFree86 is the car, not Matrox,
and Matrox is a different octane fuel. Commercial service
vehicles can run on that octane, but your XFree86 car can't.
] If matrox went bankrupt, their cards may NEVER work on future
] O/Ss! Paradise, and others, could go bankrupt, and their
] cards are almost ASSURED of working in ANY future O/S!
Is this the sole basis of your purchase decision? Then tell
them, not me. The will not have to release specs to satisfy
your "whati if Matrox went bankrupt?" scenario, however: a
source escrow would be sufficient. They probably already have
one if they have a GSA number to allow for government sales.
] VM86 (virtual machine 8086), IS basically real mode. GRANTED, it
] maps new memory into the conventional mode area, etc, but that
] is it. This is also used by dosemu. It is NOT, however,
] efficient, and many are even talking about doing away with it.
] Do you REALLY think someone is going to buy a card that is
] supposedly so fast on video, and will be made happy by slowing
] down THEIR WHOLE SYSTEM by doing what you suggested? Please,
] let me know who. I'd like to sell them a bridge!
This is different from your claim of VGA incompatability. Now
you are claiming a speed differential -- so what? The speed
differential is what you pay for. Part of it is the price
of the card itself and part of it is the price of licensing
proprietary (non soruce) software for it. Again -- so what?
] >I'd like to see you document their claims to register level
] >compatability. The VGA standards only documents INT 10
] >interfaces (unless you are mistaking the IBM implementation
] >documentation for a standards document?).
]
] Whether it is called a standard or not is really pretty
] meaningless. It IS a de-facto standard, etc.... I don't
] know if the original LRM for C had stdio, etc... By the
] time they became widely used, etc....., it was ASSUMED that
] C WOULD have it! The FACT is that almost all cards DO!
"ASSUMED" and "complies with VGA standards" are two different
arguments.
If you personally assume that "VGA" means things other than
what it is documented as meaning, well, then, I guess you'll
eventaully pay the price for such an assinine assumption.
] NO, I am talking about a COMMON solution that would, and HAS,
] worked in EVERY case! Since it has worked in every one of
] millions of cases so far, what is different about this one?
NO, I am talking about a FLATBLADE screwdriver that would, and HAS,
worked in EVERY case! Since it has worked with every one of the
milliongs of flat blade screws so far, what is different about
phillips head screws?
Do you see my point? It is stupid to assume outside of the
area covered by a documented standard.
] >] It ISN'T about linux, or bsd, but ANY non real mode program.
] >
] >For instance?
]
] How about windows NT before there were drivers? How about BSD?
] How about SCO? How about DOS BASED CAD programs?
Works fine as a standard VGA card, as long as I only select modes
supported by VGA using the mode select interface (INT 10) supported
by VGA, and draw to memory (which is mapped where VGA requires it).
The only overhead above and beyond the ken of normal protected
mode drivers is the mode switch. And I argue that mode switch
code is "boot code" -- unworthy of optimization, with no
proportional benefit from pursuing optimization.
] >I remind you that "ANY non real mode programmer" can get
] >documentation as long as he does not disclose it by publishing
] >source code.
]
] And have routines for each of HUNDREDS of cards? NO THANKS!
Your choice; be a real-mode programmer then, or use a supported
interface as a building block, but don't bitch about your
preferred foundation not being supported when you are unwilling
to take up the torch and performe the required work to convince
Matrox that their policy does not provide additional legal
protection and should therefore be changed.
Screaming at a barrier will not remove it; convincing the owner
that it is not to his benefit to have a barrier there, will.
Terry Lambert
terry@lambert.org
---
Any opinions in this posting are my own and not those of my present
or previous employers.