home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.os.os2.advocacy
- Path: sparky!uunet!microsoft!hexnut!bobatk
- From: bobatk@microsoft.com (Bob Atkinson)
- Subject: Re: FCC will proclaim Microsoft is run by Communists! : )
- Message-ID: <1993Jan05.030334.16877@microsoft.com>
- Date: 05 Jan 93 03:03:34 GMT
- Organization: Microsoft Corporation
- References: <1993Jan2.004712.10346@spang.Camosun.BC.CA> <1993Jan02.091939.18120@microsoft.com> <8292@lib.tmc.edu>
- Lines: 32
-
- Jay Maynard writes:
- >> If you claim to be 100% Intel compatible, and your processor
- >>exhibits difference in behavior from Intel's, should Intel be
- >>responsible for changing their product?
- >
- >If you claim to be 100% Intel compatible, and your processor exhibits
- >identical performance to an Intel processor's documented behavior, you _are_
- >Intel compatible.
-
-
- What if 10s of 1000s of Intels customers, or a small number of
- Intel's high-volume customers, choose to rely on implementation details
- that were not part of the "documented behaviour?"
-
- I think the most useful definition of "compatible" is the pragmatic one: is
- it substitutable by customers, not a theoretic one.
-
- >> You bring up functional specifications below - the fact is,
- >>MS-DOS doesn't have a rigid one that would allow a clone
- >>to be written.
- >
- >By default, then, MS-DOS' published API qualifies, especially since MS will
- >tell you every time you turn around that that's all a programmer should
- >use...never mind that MS breaks that rule regularly...
-
- I'll comment in analogy about the published Windows 3.0 API, since that's
- one I know more about than MSDOS. I know I was shocked when I learned about
- the amazing amount of third-party app testing that went into Win3.1. It's
- very very clear that production applications rely on much more than the
- published documentation. Message orderring, for but one very small example.
-
- Bob
-