Meld roadmap proposals

[Kai Willadsen <kai willadsen gmail com>]

Hi all,

Following the first Meld release in a long, long while, I'd like to
propose a possible plan for upcoming releases. We are currently pretty
generous with our support for old versions of dependencies, requiring
only Python 2.4, Gtk+ 2.8, and some form of libglade and GnomeVFS.
However, this generosity is probably going to start to hurt with the
upcoming Gnome/Gtk+ 3 world, not to mention Python 3 and PyGI.

I'd like to propose a plan for moving Meld to newer dependencies in
three stages:
 * Merge any interesting outstanding work/bug fixes and move towards a
new stable release series - call this Meld 1.4.
 * Port Meld to slightly-newer equivalents of Python/Gtk+/GLib (maybe
GtkSourceviewer as well) and have a fairly short unstable release
series culminating in a stable release series targeted at the
Python/Gtk 2.X series - call this Meld 1.6.
 * Move to the new-world-order of Python 3, Gtk+ 3, GLib 3, PyGI etc.
- call this Meld 2.

The first part of this could happen very quickly - within a month or
two. For the second part of this, we need dependency bumps. It's hard
to know exactly what we'll require, but as a first draft:
 * replace libglade with GtkBuilder (Gtk+ 2.12, but really we need
2.14 for add_objects_from_file)
 * move from GnomeVFS (and native Python IO?) to GIO (at least GLib 2.16)
 * bump our Python dependency in step with these, either to 2.5
(probably shipping with the required Gtk+/Glib versions) or possibly
2.6 (for helping port to Python 3)

As to the third part, well... anyone want to help?

Anyway, this is just a proposal at this point. Thoughts, feedback,
etc. is welcomed.

cheers,
Kai