Subject: SourceForge (was Re: impolitic statements (business model clues for the clueless))
From: Tom Lord <>
Date: Wed, 17 Oct 2001 19:45:23 -0700 (PDT)

       SourceForge is one of the best assets in the community. 

Undoubtedly.  And I do believe lma's pitch -- that's its a good tool
for reducing certain kinds of undesirable entropy, especially in out
of control organizations:

	Today it's a challenge for senior management of a typical
	large company to even list all of the software development
	projects going on in the company, much less track the status
	and resources assigned to each project.

A dismal state of affairs.  Valinux deserves success in this area for
identifying the problem and presenting the market with a tractable

But I'm a nit-picker and have some strategic directions for FSBs in
mind, so, sorry, I do have to gripe about SourceForge.

There's a kind of inevitable "observer effect" in tools like SF.  They
not only catalog engineering activity -- they must also, to an extent,
shape it.  SF got first to market with a solution that was ripe, given
the technology and emerging popular practices as it was initially
being developed.  In its details, it can be radically improved.

It's an issue I consider important because tools which hit the most
abstract bullet points of SF are going to be important as FSBs grow in
the long term.  "Open source practices" has to grow very far beyond
the slogan stage to become a deeply significant description of the
breeding ground for Free Software.  It has to come to mean a
collection of engineering practices that are more than just the common
glue of some standard mailing lists, remote CVS, and a menu of
project-status cataloging options.  It has to come to mean much more
than just a handful of cults-of-personality around the maintainers of
the most widely used packages.

I hate to keep holding up MSFT as the spectre of our doom, but
*there's* an organization that by all outward appearances is highly
coherent -- one gets the distinct impression that there, teams of
high-level, well-qualified decision makers develop and continuously
refine comprehensive software strategies, then develop tactical plays
out of that strategic vision.  One also gets the impression that they
are making progressing in leaps and bounds in terms of their ability
to execute those tactical plays. (And it's interesting to compare that
situation to the challenges described in the above quote from lma.)

We in the FSB world now face the additional challenge of creating
similar coherence and accuracy *across corporate boundaries* in a
situation where several of the key players, like their proprietary
Unix progenitors, are busily competing with one another over minor