Re: g-wrap parallel installability part deux

[Andreas Rottmann]

Greg Troxel <address@hidden> writes:

>   Also, I wonder why the wrapper is so important - are that many of our
>   users (who install the new G-Wrap to build guile-gnome) going to also
>   build GnuCash?
>
> IMHO, if the new g-wrap doesn't completely supercede the old - work
> with all old sources, and old glib -- and this seems to be the case --
> then we need parallel installation, perhaps by as simply as calling it
> gwrap2 everywhere as its name (directories, libs, etc.) instead of
> gwrap.
>
Yes, it's true that the new g-wrap isn't a drop-in replacement, even
with the compat layer, which is pretty complete now (I was able to
build GnuCash with it), since it doesn't support GLib 1.2. I'm
thinking about changing that and ship a GLib 1.2 wrapper, but I'm
somewhat hesistant due to the code duplication this involves.

> Giving up on classes of users doesn't seem like a good strategy.  If
> I wanted to run gnucash, I'd build it from source using netbsd
> pkgsrc.  This would build the old g-wrap and install it.
>
Hmm, can't do much about that; if I get the new g-wrap to support GLib
1.2, maybe the netbsd pkgsrc should be updated.

> I've been wanting to play with the new guile-gobject for a while, but
> between a dizzying array of branches and arch repositories and g-wrap
> issues, I've not tried that hard.
>
Well, the arch thing shouldn't keep you from trying - there are
recent tarballs of both G-Wrap and guile-gnome available.

Andy
-- 
Andreas Rottmann         | address@hidden      | address@hidden | address@hidden
http://yi.org/rotty      | GnuPG Key: http://yi.org/rotty/gpg.asc
Fingerprint              | DFB4 4EB4 78A4 5EEE 6219  F228 F92F CFC5 01FD 5B62

Software Patents: Where do you want to stifle inovation today?