Re: [Axiom-developer] Which source distributions should we list?

[Ralf Hemmecke]

> Axiom.build-improvements and Axiom.wh-sandbox will deal with the
> manufaturing of necessary patches as they can.

What do we have on Sourceforge?

trunk this was some snapshot of the
build-improvements   branched from trunk
wh-sandbox           branched from build-improvements (r263)
gdr-sandbox          branched from build-improvements (r335)
daly                 branched from trunk (r6)

woodpecker:~>svn log -v https://axiom.svn.sourceforge.net/svnroot/axiom -r6
------------------------------------------------------------------------
r6 | (no author) | 2005-01-05 00:59:09 +0100 (Wed, 05 Jan 2005) | 1 line
Changed paths:
   A /branches/daly (from /trunk:5)
   D /branches/daly/CVSROOT

This commit was manufactured by cvs2svn to create branch 'daly'.
------------------------------------------------------------------------

So to automate things, it would be best if Tim gets his branch in sync 
with trunk and trunk will become the main line.

The first (automated) merge of build-improvement back to trunk will 
probably be terribly hard, because Tim did not track which (svn) 
revisions he took from Gaby's and Waldek's branch to apply them to the 
daly branch.

Since silver is just a copy of the tla archive axiom--main--1--patch-50, 
I think it can be completely removed. Tim is now trying to make daly to 
the silver branch.

I still don't see a good structure. And I actually will not open up any 
new branch until I finally know what the mainline is and how the 
development goes.

I somehow don't like to create and sent patches. Why do we have version 
control systems then? Even if SVN is controversial, what is so hard to 
let the trunk as sourceforge play the role of the Axiom mainline?
(Gaby and Waldek are currently doing almost all of the commits.)

Yes, it's the old non-distributed development. But the Axiom team has no 
Linus. In fact, kernel development centers around Linus, right? So there 
is no very big difference at the moment for the 3 or 4 active Axiom 
developers (who actually commit) to work centralized or distributed.

We can change that later if we have thousands of contributors. But until 
then we should do more to attract developers.

Ralf