Comment 14 for bug 679715

Hi, I haven't tested this, but just wanted to add that this note about gettext domains/catalogues.

With linux packaging it makes sense to split the tools that use wx from the tools that don't e.g. on fedora the 'hugin-base' package can be installed independently of the 'hugin' package. So it would make sense that if all the tools are to be translatable then there should be two domains (i.e. two .pot files).

i.e. these are all the tools that use wx and that would be in the 'hugin' domain:

 /usr/bin/PTBatcher
 /usr/bin/PTBatcherGUI
 /usr/bin/hugin
 /usr/bin/hugin_stitch_project
 /usr/bin/icpfind
 /usr/bin/nona_gui

...these are the tools that don't use wx and which should be in a separate 'hugin-base' domain:

 /usr/bin/align_image_stack
 /usr/bin/autooptimiser
 /usr/bin/calibrate_lens
 /usr/bin/celeste_standalone
 /usr/bin/checkpto
 /usr/bin/cpclean
 /usr/bin/cpfind
 /usr/bin/deghosting_mask
 /usr/bin/fulla
 /usr/bin/hugin_hdrmerge
 /usr/bin/keypoints
 /usr/bin/matchpoint
 /usr/bin/nona
 /usr/bin/pano_modify
 /usr/bin/pano_trafo
 /usr/bin/pto2mk
 /usr/bin/pto_merge
 /usr/bin/tca_correct
 /usr/bin/vig_optimize

This has an advantage in that translators can choose to just work on the GUI tools, or they can choose to work on both catalogues.