msn-pecan 0.1 good enough?

[Felipe Contreras]

Hi,

Back in 2008 there was a proposal to use msn-pecan, and even though I
wasn't sure about msn-pecan's stability, I promised to do my best to
fix as many bugs as possible:
http://adium.im/pipermail/devel_adium.im/2008-June/005183.html

However, Pidgin's MSNP15 came into play and the decision was to switch
to it, leaving open the possibility to switch back to msn-pecan once
the features were outgrown:
http://adium.im/pipermail/devel_adium.im/2008-September/005765.html

I believe this moment has come with the 0.1 release which I consider
to be rock-solid, and has the most requested features: direct file
transfers, and offline messages (sending and receiving). Pidgin
doesn't support direct file transfers, and probably won't any time
soon.

I hope at this point in time it's clear which protocol plug-in is
better. However, there are more important reasons why I think
msn-pecan should be used, and my argument resides in three premises:

1) I am the one that has better knowledge of *both* code-bases;
msn-pecan, and stock libpurple

As you can see this an old blog post [1]; 42% of my code from 2004
hasn't been changed. The second biggest contributor is QuLogic with
only 18% (way behind). If you use stock libpurple you'll be trusting
two guys that wrote only 25% of the code.

2) Pidgin guys have admitted the MSN protocol is *under-maintained*

John Bailey explained in his blog[2] the reason of their negligence
regarding the MSN protocol; most of the developers don't care. He also
explained that they need help which is no surprise due to their lack
of expertise on their own code.

3) The plug-in is not only under-maintained, but also badly maintained

I plotted some bug statistics[3] and the results are crystal clear:
msn-pecan has fixed 78% of the valid bugs reported, while Pidgin only
37%. Even if we concentrate only on the bugs that are open at the
moment (which are in the 2 week window before they are automatically
closed), those are not properly prioritized, nor categorized like in
msn-pecan. So in essence, bugs reported to msn-pecan have much higher
chances of actually be fixed.


All in all, I don't see any future in libpurple's stock MSN plugin,
and I don't think Adium should stick with it, specially since it's the
most popular service[4]. Besides, msn-pecan does have a plan forward
[5], while Pidgin doesn't. And finally, if you find any problems with
msn-pecan, they will be tackled eventually for sure.

Cheers.

[1] http://felipec.wordpress.com/2009/08/05/who-wrote-pidgins-msn-not-who-you-think/
[2] http://theflamingbanker.blogspot.com/2010/01/on-subject-of-bugs-or-help-wanted-and.html
[3] http://felipec.wordpress.com/2010/02/05/pidgin-vs-msn-pecan-bug-numbers-dont-lie/
[4] http://adium.im/sparkle/#IMServicesWeighted
[5] http://code.google.com/p/msn-pecan/wiki/ToDo

-- 
Felipe Contreras