On Thu, 2 Feb 2006 11:46:52 -0500 Bill Nottingham <notting@xxxxxxxxxx> wrote: > Matthias Saou (thias@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx) said: > > I'm not sure I understood everything perfectly, but I'd have to agree > > about the less-than-optimal splitting, as AFAICS most of those language > > packages are 8kB big... It is not an question of size here, it is a question of being able to select the languages for which input maps get installed for m17n-db. Eventually we thinking now to have a scheme to allow the selection what input methods get enabled when an input method gets installed under SCIM, but until that is available I believe the current packaging is the best that can be done. > I expect it's not covered in comps right either - Jens, what's the > logic behind this? Right various languages should probably be added to comps: this will certainly happen for Indian scripts at least which I'm currently working on. On the other hand of course not every language needs to be comps either: I don't think most Swedes say need to have m17n-db-swedish installed for example. Jens -- fedora-devel-list mailing list fedora-devel-list@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/fedora-devel-list
