Subject: Re: Dealing with the Open Source community
From: "Tim O'Reilly" <tim@oreilly.com>
Date: Tue, 21 Nov 2000 01:48:03 -0800


kmself@ix.netcom.com wrote:
> 
> on Sun, Nov 19, 2000 at 01:27:14AM +0000, Simon Cozens (simon@cozens.net) wrote:
> 
> > We can't really have it both ways. We can't essentially claim that
> > giving a load of old code to the community is useless. Because yes,
> > OK, it's a pig if corporations dump a load of abandonware code onto
> > the free software community; but, we *do* ultimately reap the benefits
> > of this because there is always some coder somewhere who is
> > intelligent enough to make use of it. We took the HTML rendering code
> > from Netscape and turned it into a library that we can build things
> > like Galeon on top of; we'll take the file format stuff from
> > OpenOffice and turn it into convertors - I'm sure Sun didn't expect
> > this, but it's what we, as a community, need to have.
> 
> Violent agreement, Simon.  I'm not saying that these projects are
> without merit.  I *am* strongly suggesting that the community may well
> find the parts of these projects which it likes, and which suit its
> needs (if we can talk about The Community as having One Mind), leaving
> the rest of the cruft in the dust.
>

Brian Behlendorf and I both made this point rather forcefully to the Sun
team when they called us in for consultation prior to the decision to
open source Star Office (now OpenOffice).  I made the analogy to the
breakup value of a company, in which the pieces are often worth more
separately than the whole.  In particular, we argued that they should
focus on making those filters easily available to other open source
projects, because they were probably the most valuable part.  Overall,
it's better that they open source all the code (as they did) but even
so, they should focus their efforts on making the most useful bits as
available to other projects.

That being said, I don't think that Sun thinks of OpenOffice as
abandonware, just as Netscape didn't think of Mozilla as abandonware. 
Both have continued to invest significant resources in their projects.
OpenOffice will definitely need to reinvent itself in a new release for
it to amount to much in the marketplace, pieces of it are already
useful.  And in the case of Netscape, which is further along, it's very
clear that the product produced for the corporation's benefit (i.e. the
Netscape 6.0 browser) seems fairly uninteresting, while some of the
projects that have picked up on the pieces (including not just Galeon
and skipstone, but also ActiveState's Komodo development environment and
Eazel's Nautilus) are looking increasingly important.  So the
benefits to the open source community are huge.

In this regard (off topic), I'm mindful of an absolutely wonderful
trilogy of historical novels that I read when I was a kid and that made
a lifetime impression on me, Edith Pargeter's The Heaven Tree, The Green
Branch, and The Scarlet Seed, which describe the building and subsequent
destruction of a cathedral on the Welsh border in the 12th century.  The
message remained powerfully with me, that a creative act survives the
destruction of its original container, and can inspire further creative
acts by others even as it is itself destroyed.

At any rate, I do think that it's great to encourage companies to
release even "abandonware" as open source, because it provides more raw
materials for the next generation of products.  I sometimes like to use
gardening analogies.  Think of all this stuff as compost that enriches
the soil.  If you don't compost, you end up needing to add artificial
fertilizer...

-- 
Tim O'Reilly @ O'Reilly & Associates, Inc.
101 Morris Street, Sebastopol, CA 95472
+1 707-829-0515, FAX +1 707-829-0104
tim@oreilly.com, http://www.oreilly.com