Subject: Re: the GCC steering committee
From: "Jonathan S. Shapiro" <shap@eros-os.org>
Date: Fri, 15 Mar 2002 14:08:53 -0500

> "The GNU world" will, almost certainly, be unambiguously large enough
> within the next few years

That seems to me unlikely. What I think *is* likely is that the *open
source* world is growing steadily and is arriving at the point where
coordination issues are becoming important. This is a natural consequence of
market shakeout. There are now a couple of 400 pound gorillas in the open
source world (apologies to RedHat, but could you pass the little guy a
banana please :-). This is creating stress between people's misperceptions
of equity in the open source world and the reality that "he who leads,
controls".

The latter is the essential result of open source, not the former. We need
to come to terms with that and quit pretending that the disconnect between
the incentive system and the developer social goals can be ignored. GPL
incentivizes leadership and user interests. It incentivizes developer
interests only in the sense of allowing code reuse. It does *not* (and
probably cannot) incentivize integration of changes. Doing that costs real,
ongoing investment.

As to whether the ultimate open source win lies in UNIX, my opinion is that
this is unlikely. The best "grand win" for open source is likely to arise
from disruptive change. The huge force motivating disruptive change in the
systems world over the next 5 years is security. For technical reasons, I
don't believe Windows will survive, and I don't believe UNIX or Linux will
survive. Those of you who think otherwise, note that Microsoft's "Ten
Immutable Laws of Security" got 5 wrong and 1 misfocused.

I *do* believe that RedHat will survive, because they have deep enough
pockets to adapt. Many of the tool builders and middleware builders will
survive. Most of the middleware builders won't have the capital necessary to
retool.

Clearly, in this last view, I have a bias. I may be wrong. I do think it's
worth a ponder.


Jonathan