[Adium-devl] MSN: pecan vs libpurple

[Felipe Contreras]

On Tue, Jul 15, 2008 at 7:15 PM, Evan Schoenberg <evan.s at dreskin.net> wrote:
>
> On Jul 15, 2008, at 12:06 PM, Eric Richie wrote:
>
>> Things are starting to get a bit crazy so I think it's finally time to
>> weigh in.  1.3 needs to ship.  That's the bottom line.  We committed to the
>> switch to pecan and at this point we need to stick with it.
>
> I agree, Eric. Note that 'committed to the switch to pecan' is for 1.3.  It
> is clearly the superior library to msnp9, which we were using in 1.2.x;
> there is no argument about that.  It is an iterative improvement -- with
> some significant leaps, notably personal message support and upcoming
> offline message readonly support -- and we have tested it through Adium
> betas with overall success.
>
> It would definitely hold up our own testing to switch to msnp15 at this
> point in the 1.3 beta process, though I do understand that it has been
> explored by a fair number of users through the msnp15 Adium builds and found
> to also be quite stable and overall excellent.
>
> In general, it is preferable that our libpurple codebase be that of the most
> recent release off of im.pidgin.pidgin, as closely as possible, and this is
> a heavy weight in favor of switching to msnp15 for 1.4 at the latest (and
> possibly for a well-tested 1.3.x release).
>
> Long term, it would be best if we didn't have to pick between two
> 'competing' codebases, each with its own advantages and disadvantages, and I
> think that this is clear to all involved.  Given that, I would really like
> to help (at a social/team level, not at a coding level) negotiate
> reintegration of msn-pecan's improvements into msnp15.  This shouldn't be
> impossible given that they have a common progenitor; the primary problem
> seems to be philosophical differences.

I'm not so sure the problem is philosophical differences. For example,
take a look at pidgin ticket #5762, you'll see that there's nothing
wrong with the patches, and actually there's no comment on why the
patch isn't merged. After two months I don't even get a reason about
why it's not merged.

I have the feeling, however strange it might seem, that just because I
did, there should be extra skepticism. I won't keep on trying.

So I tried to contribute directly, and that didn't work out. I'm still
making msn-pecan extremely modular, so they could just take pieces and
merge them into their code, but I doubt any of them cares enough to
actually take the time to do it.

> I know that in various places on pidgin-devl Felipe has expressed his
> frustration with various pidgin practices, and there have been heated
> arguments are both sides.  These differences may therefore not be
> resolvable, in which case it is a pity that someone's work will see less of
> the light of day than it would otherwise.

It's sad that they don't see the technical merit, yes, but msn-pecan
will be a standalone library, so it might even see more light.

Best regards.

-- 
Felipe Contreras