*sigh*
Hamish, I'm sure you mean well, even with what looks like a bogus return
address[1], but please, do a bit of homework if you actually want to help
here...
I'm sorry, but it is hard to take any suggestion for moving forward
seriously when it comes from a view that quite plainly neither knows
where we are, which direction we are facing, nor the strength or
direction of underlying currents.
So I'll point out a few landmarks for you, but I'm repeating myself here,
and that _clearly_ isn't helpful in the sense we want to see things
get helped, ok? [2]
perhaps we could establish a SVN acct with Alioth* for the package files?
Now ask what an absolutely invaluable asset that has already proved
to be for this project over the last few years ...
Then instead of half a dozen or so people making their own unofficial
diverging packages we could focus the effort in one place and all help
to prepare a really good wx2.8 package,
That's something you'll have to take up with the dozen or so diverging
package maintainers. All the people who've contacted me to date have
been given access to whatever resources they required. So this clearly
isn't the bottleneck either.
which could eventually make it into the Experimental repo and then
one day to sid?
There is a little reef the locals call 'transition plan', that wrecks
many hulls of the unwary diverging package maintainers sailing casually
between those points...
You'll have to demonstrate how to navigate that safely if you want to
get me on board.
The 'one day' you speak of is called Lenny. When Lenny arrives the
diverging package maintainers will face their final day of judgement.
srsly.
FWIW, wx2.8 release history:
2.8.6 2007-10-01
2.8.5 2007-09-10
2.8.4 2007-05-18
2.8.3 2007-03-24
2.8.2 2007-03-13
2.8.0 2006-12-12
We are now almost a year from initial release and 4-5 stabilization
releases into the cycle.
That's a pretty abbreviated history you have there.
The numbers you show above could be coup dates plotting the stabilization
cycle of governance in a third world republic, for all they show about
the release-worthiness or desirability of each wx2.8 snapshot ...
The 'time since last public mistake' isn't a great compass for future
direction. What we must care about now is what will the world-view be
when Lenny comes.
Given the last 'stabilization release' apparently requires all binaries
to be rebuilt to incorporate some of its 'stabilising', I'd be hesitant
to declare this rockpile has settled just yet.
Do you have some graph points on the number of serious bugs reported
between each of those 'stabilization' releases, and the number that
still remain? Or how many more 'stabilization releases' are projected
to be needed before this stable release is really stable enough to
carry the remaining bugs without stumbling?
Do you know if that will ever happen before:
They are now talking about Looking forward to wxWidgets 3.0
http://wxwidgets.blogspot.com/2007/11/looking-forward-to-wxwidgets-3.html
Indeed, it's actually supposed to be released before Lenny ...
And will require sourceful changes to most if not all existing
wx apps to transition from existing versions.
Seems to me that it would be a good idea to focus on 2.8 for Lenny,
... so I'm not really sure why this seems like a good idea to you.
Do you really want 4 separate wx versions incorporated in Lenny?
Or should we perhaps do some more detailed analysis of whether
wx3.x is actually the 'forward' our transition plan must deal with,
and if 2.8 is just a port with cholera which we might not visit for
this next trip.
get the package solid with as much lead testing time as possible before
Lenny's release.
Given that the remaining wx2.4 apps may have a better time switching
to wx3 than they ever did switching to 2.6/2.8, and that apps which
currently need 2.8 will drop it like a hot brick the instant wx3 is
viable -- don't you think all that energy, if you can harness it,
would not be better directed at making wx3 a more viable candidate
than wx2.8 has ever been to date?
You aren't seriously proposing two wx transitions in the time scale
of Lenny, right? I'll be delighted if we can just pull off one well,
and even that seems like an under-resourced long shot at present.
Rushing for a near-sighted goal isn't the option I'd put my betting
money on for a Good Result.
Is there support for a collaborative Alioth/SVN aproach?
This thing you call 'support', is it something far away and exotic?
Up to now, joining the group has been as easy as making a few promises
with a lot of latitude for late delivery.
But if you mean 'have _any_ of them been delivered?' ...
... then that's something you'll also have to talk to other people
about.
Or you could read what I've already said in response to this bug
previously, answering exactly this question. Which in summary is,
anyone who has the skill and inclination to do more than ask me
questions I've already
