Re: Porting to GNOME2 code - work assignment.

[Emmanuel <e allaud wanadoo fr>]

On 20.02.2002 21:24 Pawel Salek wrote:
> Hello,
> 
> I have just commited initial GNOME2 porting patches to BALSA_2 branch 
> (check it out by cvs co -r BALSA_2 or cvs update -r BALSA_2). The patch 
> is based on Jelmer's work. This is very initial stage: configure should 
> run, and one file in libbalsa should compile. The porting work is most 
> likely going to be _mostly_ straight-forward (with some exceptions). I 
> guess it should be possible to distribute the work on the file-by-file 
> basis. I.e. one might take an arbitrary source file from balsa and try
> 
> make source.o
> 
> There will be some collisions in the header files - I have already 
> noticed that signal marshallers accept different sets of arguments - but 
> I think it should be easy to resolve.
> 
> We have now already three branches:
> HEAD - where the new features come.
> BALSA_2 - porting work to GNOME2.
> BALSA_1_2_x - old but stable branch.
> 
> I do not know whether it is worth to keep BALSA_1_2_x branch alive. If 
> there was somebody willing to, for example, port some bugfixes/simple 
> features (like Reply To group) back to BALSA_1_2_x, the patches would be 
> welcome.
> 

AFAICS 1.3.x is pretty stable and yet adds new features to 1.2.x. So why 
not release it shortly as a 1.4.0-pre1 to release a 1.4.x branch soon?
	The idea behind is that it seems like we will spend a lot of time 
on porting to gnome2 that will be broken for quite a long time. Moreover 
the only feature that is really lacking is GnuPG/PGP support, but I'm not 
sure we can do both porting and adding this feature.
	I also try to begin a backport of bug fixes from CVS to 1.2.x, but 
because 1.3.x seems actually stable, I don't it is really smart to spend 
time on backporting fixes, and I think 1.2.x really lacks the 
search/filters features.
So to be short the aim is to have a more feature complete stable branch 
(1.4.x), not loose time on backport to a too behind actual stable branch 
(1.2.x), and focus on a 1.9.x devel branch which will be the port to 
gnome2.
Moreover we can also imagine that if someone is interested in implementing 
GnuPG/PGP support, he can manage to maintain a patch against stable branch 
(or open a 1.5.x devel cvs branch) because I think that it is quite 
orthogonal to changes necessary for the port (but I could be wrong).
OK, thank you if you manage to read until here ;-) What do you think of 
these ideas?
Bye
Manu