Re: Is wx salvageable? (was Re: Is Ron Lee MIA?)
On Fri, May 25, 2007 at 07:02:26PM +0200, Adeodato Simó wrote:
> Ron, while I can see how your propoosed model could improve the quality
> and usefulness of wxWidgets for the users of the library, it does not
> address the original problem raised in this thread, which is: "Stable
> releases of wxWidgets get uploaded to Debian very late".
You seem to miss the point. "stable" releases of wx tend to not
actually be usable for everyone for quite some time, and this is
a trend that is not improving with age.
Perhaps "tend" is the wrong word here. Let me clarify: NEVER has
the first "stable" release of some new wx branch gone out the door
without some critical flaw that required an immediate errata release.
It's usually not until a branch is abandoned by most of the upstream
developers that it actually becomes stable, and usable for a majority
of application developers. Especially those trying to make and
maintain stable releases of their own. That, by definition, is a
We can't just keep adding more and more wx versions to the distro,
we need a clear plan to migrate from one to the next. Without that
we will create an awful, confusing, bloated mess for users. The
new release has to prove itself at least as usable as the old one
before it can be considered a viable replacement.
> If you want to work on improving the development process of wxWidgets in
> order to produce better releases, that's perfectly fine, but once the
> wxWidgets project releases a certain version as "stable", your wishes
> for a better development process should not prevent you from doing your
> task as a Debian maintainer. If something is released, it is a
> reasonable expectation that the maintainer will upload it in a timely
> manner, unless there's evidence that it's utterly broken -- which it
> doesn't seem to be the case here.
You seem to also not read the daily influx of bug reports to the wx-dev
list. Nor offer any proof that we can transition to this release
without major breakage in an important application. That it Works For
You (perhaps) for some unspecified purposes may be nice information,
now what about for everyone else?
> Since you reckon that actually maintaining the package is not difficult,
> yet you seem to be more motivated by upstream-related stuff than mere
> maintenance, maybe you could benefit after all from some motivated
> co-maintainers? And hey, this is completely orthogonal to your quest for
> involvement from co-dependents, for which I wish you luck.
Are you volunteering to help, or demanding that I put your wishes above
carefully selecting what should be a part of the distribution and
deciding when the right time to initiate a transition might be?
As I've already said, people who have the skill and inclination to help
are eagerly invited to contact me privately about how they might do that
best. And I think the people who can help best are those who are actually
developing and testing applications that use it. People who just want to
throw the weight of their unsubstantiated opinion and personal demands
into the ring are invited to first read the constitution and social
I'll not entertain a flame war here on this topic. If you are not
prepared and capable of actually helping, then I'm afraid you are
simply not, well, helping...