[Gegl-developer] Porting GEGL

[scl <scl gplus gmail com>]

Hi,

as a foundation for the further GEGL porting work, the LGM and GSOC 2013 
I updated the GEGL porting matrix in our developer wiki [1].
Please review.

I think - despite of all the respectable work done - it's no secrets to 
say that the GEGL port has been a topic for long, is still a show 
stopper for version 2.10 and all further work and yet there's still a 
lot to do. Beside others currently 43% of the filters are ported.
Don't get me wrong - I'm not blaming anybody. Let's see how we can boost 
the process.

I think the following could help:
- Get rid of filters that don't match to the product vision or are 
outdated. Priorize the remaining filters. Alexandre wrote a mindstorm 
article [2] in the wiki and I fully agree with it. The LGM is a good 
opportunity to discuss it and also we have the mailing lists and IRC. 
Ask the artists at LGM and in our mailing lists about their needs.
- Focus the 2013 GSOC projects on the GEGL port.
- Port only a set of core functions and let the other functions do their 
work through PDB-API calls from a scripting language (Script-Fu or 
Python), if that can be done easier than writing C code. So the other 
functions could inherit silently all the GEGL capabilities.
- If other open source projects, like Darktable, Digikam, OpenCV, 
RawTherapee etc., have solutions for our missing filters, why not call 
their libraries through GEGL in a first step?

Kind regards,

Sven


[1] http://wiki.gimp.org/index.php/Hacking:Porting_filters_to_GEGL
[2] 
http://wiki.gimp.org/index.php/Mindstorm:Revamping_default_plugins_and_scripts