[Glade-devel] glade-3 maintainership

[Tristan Van Berkom <tristan van berkom gmail com> (Tristan Van Berkom)]

On Fri, 22 Oct 2004 03:11:19 -0700, Rikke D. Giles
<rgiles centurytel net> wrote:
> How did the IRC meeting go?
Hmmm, 
   lots of people are asking me that :) I'll make another post with
the log attatched.

> I couldn't attend, and I have hesitated to commit because I have little
> time for development right now.  That being said, I very much want to
> support Glade and see it move on.  I will do what I can to help with
> that.  My specialities are documentation and tutorials.  I can code
> fairly well also.

Nice to see that more and more people are interested !

We are still in heavy developments stage, hopefully we can get this
show on the road soon, there are a few chunks of code that could use
some people to take care of, off hand I would say, that some of the
editors need some work

> I'm interested in using Glade as a teaching tool.  To that end I
> consider the code generation facility, however much disdained,
> imperative.  I will probably concentrate on making some sort of module
> for Glade3 which will add this, eventually.  I am also interested in
> having some sort of modular ability to add on custom widget libraries.
> I have built myself a few of those over the years and would like to
> incorporate them into Glade.

Glade-3 features widget kits as plugins and are incorporated through
an XML catalog, any GTK+ widget kit should be easy enough to incorporate
without modifying the glade-3 code at all :-)

> I agree with all the reasons for using libglade to generate code for
> big projects.  But when you are trying to get someone into writing
> code, and do not want to teach them autogen/automake/autoconf, etc, etc
> for their first hello world, the code generation facility is very nice
> indeed.

You mean to use libglade to create a widget heirarchy, AFAIK, one of the
first and formost decisions about glade-3 is "no code generation", ommiting
this feature will allow us to focus on some higher priority areas such as
property-editing through object introspection etc... also code generation
is a pain to support, as widget kits evolve; the code generation aspects
of glade would have to be maintained, at high cost.

I've read somewhere (maybe in TODO, HACKING or on the ML somewhere)
that if people want code generation, they are encouraged to write a program
that generates code based on a glade file, I think that is a reasonable idea.

Cheers,
                                                              -Tristan