Subject: Re: Open letter to those who believe in a right to free software
From: Ian Lance Taylor <ian@airs.com>
Date: 4 Nov 1999 11:40:59 -0500

   From: "Stephen J. Turnbull" <turnbull@sk.tsukuba.ac.jp>
   Date: Fri, 5 Nov 1999 01:23:36 +0900 (JST)

   >>>>> "Ian" == Ian Lance Taylor <ian@airs.com> writes:

       Ian> The history of free software to date has been highly
       Ian> contingent.  If there were no RMS, there would not have been
       Ian> anybody else.  If there were no Linus Torvalds, there would
       Ian> not have been anybody else.

   The first is possibly true; without RMS, there would not have been
   GNU, no FSF, and who knows whether there would even have been a BSD.

   But the latter is demonstrably false.  By the time Linus wrote Linux,
   there was already Minix, there was at least one BSD, and within ten
   years there would be a HURD and three BSDs.  Linus got us here faster
   and probably at least as well as any of the alternatives, but there
   were good alternatives.

I don't believe that any of those alternatives would have seen the
general market acceptance that GNU/Linux has seen.  None of them have
welcomed contributions as Linux has, and none of them have shown any
willingness to knowingly make poor decisions in order to move very
quickly, as Linux has.  Those follow directly from Linus's character.

I think he really did introduce a new way of doing free software
development; it differs only in degree, not in kind, but the
cumulative difference after eight years is large.

   I think it highly likely that by now there would be a free OS project,
   even if there had been no RMS and no GNU.  But RMS gave us anything
   from a 10 (very optimistic) to 50 (very pessimistic) year head start.

There would be a free OS project, but it's not clear whether it would
be merely a research project, or whether it would have the explicit
goal of replacing proprietary systems.  Also, a 10 year head start is
a lifetime in the computer field.

   And heaven help us, if it started now, it would probably try to clone
   Windows NT.

That job is so hard that it would never have succeeded.  Windows is so
large that some say even Microsoft can not understand it.  Cloning it
using volunteer work, based on the inadequate Microsoft documentation,
is an impossible task.  The existing WINE project has not yet managed
to clone Windows 95, and soon they will be faced with cloning Windows
2000.

Ian