Jeremias Maerki wrote:

> On 11.02.2003 20:55:40 J.Pietschmann wrote:

...

> > Actually, the way the redesign was attempted was probably a
> > bad idea. See
> >   http://www.joelonsoftware.com/articles/fog0000000069.html
> >
> > Seems we have now the worst of all, with a reasonably working
> > but outdated code in the maintenace branch and a huge investment,
> > but still not fully working code in HEAD :-(
>
> We took a decision and I can still live with it. Nobody said it would be
> easy. Of course, it would be nice if we got some more resources in form
> of people working on the redesign or even funding to developers or
> committers... The biggest problem is probably that we lost a lot of
> knowledge (and possble workpower) when important committers left. At
> least you can say that the redesign holds some promising prespectives
> (IMO).

Sorry, I have to agree with Joerg here, although I have no desire to reopen
the issue. I would rather have done a million refactorings to get us where
we want to go than to maintain two sets of code. I think that the reason
resources are not forthcoming is because of the barriers to entry that we
have raised, the largest being that, if you want to work on the code, you
really only get to work on the code that doesn't run. For most people, that
is a no starter. I think when the trunk works reasonably well, resources
will appear again. <Caution: Blatant pontification follows> The sine qua non
of open source software is that the code works. Any experimental code that
doesn't work but needs to get checked in should go onto a branch until it
works. Commercial software developers might be able to get away with a
different approach, if they hold enough cash in their hand to buy the
resources to get them across the hump. The transitory nature of open source
resources dictates that every incremental change (on the trunk anyway) be
pretty much logically complete, and (by intention anyway) better /even in
the short term/ than what was there before.

All of that said, I am committed to the redesign, and frankly can't wait to
get back to it. This is not only an extraordinary project, but it has an
extraordinary crew (but let's don't ever do it this way again ... please).

Victor Mote


---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, email: [EMAIL PROTECTED]

Reply via email to