Release team perspective [Was: libgnomeprint* 1.116.x in 2.2]

[Jeff Waugh <jdub perkypants org>]

<quote who="Chema Celorio">

> > This sounds like a good idea to me, as we can support the use of
> > applications built against 2.0.x on 2.2.x. Even though libgnomeprint*
> > doesn't have huge user desktop or application penetration [1], it sets a
> > good precedent in case we do some crazier changes further on in 2.x.
> 
> Sounds like a good idea to me too. I wasn't too excited about changing
> the API and had expected that the transition to be a lot bumpier, but it
> ended up being ok. I'm all for it, we don't force users of gnome-print
> to have to upgrade RIGHT NOW, they can do so whenever they need a
> feature of a newer version.

So, we just had the release team meeting and discussed this. Jody had some
"very warm words of affection" for gnome-print 1.116.x, and argued that we
should strongly encourage app developers to target the 2.2 version instead -
to the point of deprecating 1.116.x with great prejudice.

  - The 2.2 version does not have any dependencies on other 2.2 modules
    (though has an optional dependency on fontconfig).

  - The 1.116.x version should simply be avoided.

  - Most major apps can target the 2.2 version without much fuss. Gnumeric
    is sitting on the fence, Evolution will probably be targeting the 2.0
    platform but can pull in gnomeprint 2.2 without much hassle, etc.

  - Though we would like to set a precedent for compatible desktop
    libraries, we only guarantee this from platform libraries, and
    gnomeprint is a special case (in that we don't want developers using
    it).

So the answer from us roughly ends up being that this time, we should not
set a precedent for desktop library compatibility, because we'd far prefer
to see the previous version disappear.

Does this sound okay?

- Jeff

-- 
  "There's always a new bogeyman - every two months, there's a new axe to   
                 add to the axis of evil." - Michael Moore