httpd-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Ben Hyde <>
Subject Re: Apache History Project - Call for comments
Date Wed, 15 May 2002 16:04:22 GMT

Thomas Eibner wrote:
 > I agree, what I was more interested in was actually the sheer volume of
 > the mailinglist archive. 

I've graphed volume (particularly in CVS's logging) over time, that's an
interesting way to keep an ear to the railroad tracks.

 > These statistics weren't created for someone to
 > be offened and they aren't going generated on a regular basis anyway.
 > This was not an attempt to degrade anyone's effort on the Apache httpd
 > list. 

I apologize if I left the impression that anybody's motives weren't of
the highest caliber.  You wouldn't believe the damage I've mindlessly
done to projects with awk.

Rich Bowen wrote:
 > .. interesting statistic, ... get peoples' attention ...

That it's fun and grabs attention is just part of the lession I learned
the hard way.

 > ... roadmap ...

My current favorite reductionist model of open source is that you have a
pool of common knowledge manifested in the repository/web-site/mailing
list archives.  The cell wall of this pool is managed by those who have
commit rights.  That 'innovations' are pushed into the pool, and
releases are pulled out.

For example CTR is a rule about the details of the cell's digestive
tract as is the you can't veto a release rule.  

Another, I think unspoken rule, is that if somebody does the work - we
strive to empower them.  That's the way I feel about the history project
- cool go for it.  That was the way I felt about the budding off of
APR; very risky but I bit my tougue.  That seems to have worked out
quite nicely.

The second part of my tiny model is that there is a cycle: releases are
refined and distributed to users who use them.  Some of those users
innovate, some of those innovations get sent 'home again' some of those
get noticed, some get pushed into the pool.  One of the oh so many
reasons to keep it open is it increases the chance of somebody out
there getting a cool innovation back into the pool.

Getting the cycle to function better, understanding how the various
stages work, where the barriers are, etc. etc. is all facinating.  For
example, cvs diff was a huge breakthru in making the flow of incomming
stuff work better - it really focuses the discussion and helps the
contributor to know what to do.  We haven't made much progress on that
front since!

In the early days this cycle was a lot tigher, simpler.  These days, as
it's scaled, we have lots of larger institutions all around the cycle.
That has benefits and agency problems.  Looking at the history of this
evolution should be fun and facinating.

Sorry... this entire discussion is off topic.  This mailing list
is were the cell wall _is_  this discussion is 10K feet over that.

Have fun.

 - ben

There is honor in the mail not written.

View raw message