[Sussex] Gentoo forked...

Steve Dobson SDobson at manh.com
Tue Jul 1 10:02:00 UTC 2003


Morning Mark

On 30 June 2003 at 21:33 Mark Harrison wrote:
> I have to disagree here.

Well it would be a very boring list if everyone agreed with me.

> I've been trying to cat-herd a bunch of developers in an 
> OpenSource project (a protocol) for the last year. We have
> about 6 developers who are KEY to development of the feature
> set.
> 
> In the year, we've formed a very good working relationship, 
> and are very open with each other, but not at all guarded.
> This has only been possible on the basis that all our
> discussions are absolutely confidential, and that we will
> always present a united front to the wider community once
> we have agreed on something.
> 
> 95% of agreement is made by discussion till we all agree on 
> something better than any of us had originally agreed. About
> 5% has come down to a simple vote.
> 
> I do not believe that we would have anything like as good 
> (technically) or useful (in terms of deployment among the
> interoperability community) if we'd just had open discussions...
> I believe this because after 3 years of open discussions, we
> were still left with NO actual products used by more than
> the author .... until we changed to a different basis.
> 
> I am not for a moment saying that the approach we took is
> universally applicable - instead I am saying that I don't
> believe that ANY approach is universally applicable.

Please, don't get me wrong.  I wasn't saying that the Debian
way is the only way.  A small group bashing away at the start
of a project is not going to be in the same public spotlight as
the big projects.  It was for these large distributed open
source comments that my comments were aimed.

Does a small, closed group come to a decision quicker: absolutely.
Does a large, open to all, group take ages to come to a decision: 
well after a lot of debate and cross talk, not to say the odd
flame, a decision can be reached. :-)  I never said that there were
not problems with the open forums.

My point was that open forum does not close itself off form any 
idea that someone is prepared to offer.  If the idea has merit then
others will take up the call.  Debian could make a dramatic change
based on an idea of someone who's only contribution to Debian is 
that one posting.  Gentoo, by limiting who can post, to a public
forum is censoring themselves.

Your "bunch of developers" forum sounds, to me, more about control
than censorship.  And in some respects why should you give up control;
it's your baby.  Control of your idea is a very difficult thing to
give up.  You're not along - Linus still has control for the main
Linux kernel source.

I turned to open source because it provided me with a vast array of
software for only the cost of returning some of my expertise to the
community.  But that is not either of the reasons that I think that
open source Development is going to win in the software market place.
It will win because:

1). Open source is more responsive to it's clients needs - just by
    allowing them to add their own features.

2). [This was not one that I foresaw] That open source tends towards
    bug free software quicker than close source projects.

History in software is being to show us that opening a project up can
provide some great advantages.  I put it to you that the more closed
a project is the more chances there are of a clone or fork starting up.
In the software environment, like nature itself, it is survival of the
fittest, and fitness for software is defined by my two point above.

Steve

[1]
http://www.sitepoint.com/article/767





More information about the Sussex mailing list