On 2016-08-28 01:51 AM, Thomas Martitz wrote:
Am 28. August 2016 01:09:36 MESZ, schrieb Matthew Brush mbrush@codebrainz.ca:
Hi All,
With respect to the efforts described in PR #1195[0] is anyone opposed to using Vala as a GObject code generator?
I propose we use Vala as a way to generate GObject boilerplate while we
hammer out the design of FT-plugins, and later once the design is more concrete, that we port the GObjects generated by valac to their hand-written C equivalents (after they won't be changed as much).
Do you suggest to use it only initially, and then continue working on the C code? In that case the vala code doesn't need to be checked in does it?
I suggest to use Vala to write any GObject stuff for prototyping purposes in order to avoid writing and re-writing gobs of boilerplate. The Vala code would have to be checked-in in order to modify it until a design is hammered out then the code could be converted to plain C boilerplate once it's not changing often (before merging back to master branch).
The classbuilder plugin (shipped with geany) can also generate the boilerplate, bit much prettier with less noise.
The classbuilder plugin only generates basic GObject boilerplate, it has no concept of interfaces, abstract base classes, properties, signals, etc.
If you suggest to continue working on the vala code that's ok with me but not everyone is fluent with vala.
It would be OK with me too, but I suspect not everyone.
Cheers, Matthew Brush