[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: wxwidgets 2.8 needs help !

On Sat, Nov 10, 2007 at 10:59:06AM +0100, Francesco P. Lovergine wrote:
> On Fri, Nov 09, 2007 at 04:03:56PM +0100, Loïc Minier wrote:
> > On Fri, Nov 09, 2007, Francesco P. Lovergine wrote:
> > > So what about that? 2.8 is around since almost 1 year, and this delay
> > > is not justifiable IMHO. There are now tons of developers and
> > > maintainers that need to build against 2.8, and I doubt a transition
> > > plan from 2.6 has sense due to API and behaviors changes. We have
> > > simply to cope with two different versions, as already happens 
> > > for other libraries.
> > 
> >  If you're busy-waiting on it, I guess you can propose to take over
> >  maintenance starting with wxwidgets2.8?  Looks like a man power issue.
> > 
> Honestly I think not. Current maintainer is discouraging 2.8 packaging,
> plain and simple, as you can see by wxwidgets 2.6 reports. I can also
> partially justify him,

So if you want to hurry this up, learn more about the issues and
help to fix them.  Saying "maintainer is discouraging 2.8" makes
it sound like you think that if the maintainer just suddenly
stopped 'discouraging it', all the reasons for doing that would
also just magically go away.  You seem to acknowledge that you
share at least some of my concerns, though you don't say which
ones ...

> but to be honest I would understand better
> if he would orphan that package instead, as many people do when
> they have to cope with brain-fucked upstreams and have neither time
> not will of doing that.

Uh, so your solution to 'there is more work to be done before this
is a viable release candidate', is for more people to toss in the
towel leaving even less people to do the work ...  ?

Are you proposing a transition plan to a future wx version, or
proposing that we ditch it from the distro completely?

While there are users prepared to do their bit to ensure the bits
they need work well enough, I think the latter is a bit premature
irrespective of how brain-fucked any particular upstream may be.

> Delaying and complaining is not a solution and do not help to
> achieve a better quality of the product.

Given that a premature release is largely the cause of the problems
in 2.8 to begin with, delaying a major transition until it has a
chance to go smoothly is actually exactly what we are worried about
from a qa perspective.  Better Quality doesn't magically appear,
it is the result of concerted effort by interested people.
Sometimes we have to just wait for that effort to be completed.

The complaining is mostly just coming from people who don't seem
to give a damn about the quality of the product, who aren't doing
anything to help clear the ~1500 upstream bugs currently still
open (a list which doesn't appear to be getting shorter), who
aren't doing anything to assess the ability of existing apps to
transition, and who can't be bothered getting experimental betas
of such software from other more appropriate sources for themselves.

> If all of us did that, we would have to drop tons of libraries
> and programs out there and probably the whole project.

Don't you just love the great rush into chaos at the end of a
good slippery slope ...  ;)

There are a number of places that people who really need 2.8 packages
of varying degrees of quality can get them from.  Chances are, all of
those people will also really need wx3 packages by the time Lenny is
frozen.  wx3 isn't ready for Debian yet, but at this stage, if you
forced me to bet hard earned money, I'd probably side with the
proposition that it will be before 2.8 ever is...

Seriously Francesco, if you actually want to help, all you have to
do is tell me which bit you want to help with, convince me that you
have a well thought out plan for it, and then do it ...

If you don't, then please don't try to tell us how we should do it,
we already have some clever people looking into that, and the answer
really is a bit more complex than 'ooh, upstream made a new snapshot,
quick chuck it in the distro' -- at least if we want our users to
keep believing it is really that easy in preference to having them
load testing the bts with the fruits of their disappointment.


Reply to: