Re: [Axiom-developer] A suggestion for Gold

[Bill Page]

On 6/26/07, William Sit wrote:
> The following email (with minor changes) was originally sent
> to Tim and he replied: "Post your suggestion publicly. Ask
> for a vote. We will go with the majority opinion. You get to
> keep the final count."
>
> So please feel free to comment, and may be Bill Page can set
> up a sandbox page to keep votes? (sorry, I don't have the
> expertise).

http://wiki.axiom-developer.org/VoteASuggestionForGold

When someone asks for a vote and it is seconded by another, William
Sitt and Tim Daly in this case, I suppose the usual "rules-of-order"
would say that we should just vote with a minimum of further
discussion. But since we seem to be making the rules as we go, I can't
resist the opportunity to make a few comments before deciding how I
should vote.

First, I want to thank William for suggesting such a bold but perhaps
obvious move. This is certainly one way to resolve a lot of problems
and move ahead very quickly without further risk of a fork in the
Axiom project.  It also has the merit of apparently being quite simple
and easy to accomplish. After building and using wh-sandbox on several
different machines I also have a lot of confidence that wh-sandbox is
a very significant improvement over the current Axiom Gold release.
But is it too big of a quantum leap?

Thanks to Waldek's admirable energy and skill wh-sandbox is still
evolving quite rapidly. Should we really just pick a current revision
of wh-sandbox and merge it with Tim's current Silver (= SVN trunk) as
William proposes? We can define the result as the new Silver, but then
what happens to build-improvments which thanks to Gaby's continuing
effort has also been evolving on a somewhat different but related
path? wh-sandbox was branched from build-improvements but that is now
several hundred revisions ago. Would this make it easier or more
difficult to merge build-improvements also back into SVN trunk?

And what about the actual process? We would presumably rather quickly
promote this new Silver to new Gold so that the largest number of
Axiom users would be able to take advantage of these changes. But are
we risking the possibility of a serious reversion of functionality if
the testing is not sufficient and if the documentation remains at it's
present minimum level?

I agree in principle that open source development should proceed at a
very aggressive pace. That is the best (only?) way by which we can
hope to keep the interest and attention of the largest number of Axiom
developers and users. And I think it is reasonable to take a fairly
high level of risk - moderated by use of modern source code management
tools that in principle make it possible to manage multiple versions
of source code very easily. Also we should acknowledge that Axiom in
it's current state is still very much a "research work in progress"
and *not* a direct competitor for the existing commercial alternatives
when it comes to "production oriented" applications. So we are
therefore free to engage "warp drive" when we feel we need to without
risking it all down some wormhole. (If you will tolerate my extended
metaphor. :-)

Of course all of this depends on the willingness of the primary
maintainers of wh-sandbox, build-improvments and Silver to co-operate
in the suggested manner. Without that, "voting" on the issue seems
rather moot.

I think I will ponder these questions of another 24 hours before I
decide my final reply to this proposal.

Regards,
Bill Page.