Re: GIR-based vala bindings

[Jiří Techet <techet gmail com>]

On Thu, Mar 31, 2011 at 10:29, Lorenzo Masini <lorenxo86 gmail com> wrote:
> Il giorno mer, 30/03/2011 alle 18.37 -0600, Robert Park ha scritto:
>> On Wed, Mar 30, 2011 at 4:16 PM, Lorenzo Masini <lorenxo86 gmail com> wrote:
>> > Robert: probabily you are the one who best tested python bindings.
>> > Since I have made some changes in gobject introspection files, could you
>> > give them a try after Jiri have committed all the stuff?
>>
>> Is it committed yet? I just rebuilt my jhbuild environment and got
>> libchamplain commit 9c42276d437eaede6337c18b50ab4dbad6eb4ed7
>> ("0.9.1"), and my test suite passes fully here.
>
> No, Jiri is reviewing it.
> If you have time you can try from my personal clone
> (https://gitorious.org/~rugginoso/libchamplain/rugginoso-clone).
> Note that it isn't aligned to master, because I was too lazy to git
> rebase :P
> The changes are in the branch vala-switch-to-gir.

The patches are in master. The vala bindings generation is _really_
nice now. I thought that the gir-based generation is possible only
with a development version of vapi generator and I was surprised it
worked fine with Vala from Natty. Good work!

I have fixed a few minor issues and updated the demo to the new API.
I've also diffed the gir's generated now and before and they only
differ in what other girs are included (you've correctly removed the
gtk dependency in champlain, which wasn't necessary). So I don't
expect the changes will have any impact on the Python code.

Cheers,

Jiri