Subject: Re: Meeting the MS challenge
From: "L. Peter Deutsch" <ghost@aladdin.com>
Date: Mon, 1 Dec 97 07:19 PST

> Microsoft is not worried about libre software volunteers.  They're
> worried about a proprietary software firm deciding to create a
> Windows-compatible system.  That's WHY Microsoft constantly extends,
> and then collapses its APIs.  And that's the chief public bad produced
> by a proprietary OS -- increased development costs for everyone else.

I agree.  Nevertheless, in order to create a libre software platform capable
of running MS Windows applications, it is still true that libre software
developers must keep up with Microsoft's API changes, and therefore the
assertion that "Microsoft is generating new extensions to Windows faster
than any conceivable horde of libre software volunteers can emulate them" is
relevant, even if it isn't Microsoft's explicit intention or primary
concern.

> Presume that existance of a version of Linux which runs all Win95
> applications.  You would have much more success persuading several large
> proprietary application vendors that their development and support costs
> would be lower by creating native Linux versions of their application.

But that goes back to my previous posting in which I point out that the
software infrastructure available in the Unix world for developers of
Windows-perceived-quality interactive applications is (1) seriously lacking
in GUI support libraries, (2) unstandardized and uncoordinated, (3) poorly
supported by tools and documentation, and (4) not progressing significantly.
Those things don't matter if one only wants to run Windows applications, but
they matter tremendously for native Linux applications.

It is true that the proprietary application evelopers would save one-time
costs of a few $K per seat, and annual costs of perhaps $1K per seat, by not
having to buy proprietary SDKs, tools, and libraries (assuming, of course,
that libre software proponents are successful in dissuading anyone from
developing Windows-native libre equivalents of these things), but I believe
these costs are very small compared to the costs implied by #1 through #4
above, and are negligible compared to the value of having applications that
run on Windows versus applications that don't.

I'm assuming that you were talking about persuading application developers
to *switch* from Windows native to Linux native.  If you were talking about
persuading application developers to *add* native Linux support, that's a
different story.  All that requires is a large enough installed Linux base
and a perception that Linux users are willing to pay the going commercial
rate for good (proprietary) software.

-- 

L. Peter Deutsch         |       Aladdin Enterprises :::: ghost@aladdin.com
203 Santa Margarita Ave. | tel. +1-650-322-0103 (AM only); fax +1-650-322-1734
Menlo Park, CA 94025     |        http://www.cs.wisc.edu/~ghost/index.html
*Oppose bulk-mail abuse of your mailbox and newsgroups: http://spam.abuse.net*