On Wed, Apr 28, 2004 at 07:44:23PM +0200, Henning Meier-Geinitz wrote: > Hi, > > On Wed, Apr 28, 2004 at 03:08:02PM +1000, Matthew Duggan wrote: > > Worse, some backends initially define values as SANE_I18N, then do > > strcmps in other places without the SANE_I18N decoration. This will > > surely cause trouble for people using translated versions of SANE. > > As Mattias already daid: this is not a problem. If it had been a problem, > SANE users would have cried out loudly years ago :-)
Ahh goodo, I have to plead ignorance when it comes to the mechanics of internationalisation stuff. ... > > outside of en.US for that matter). This is also possibly a bug in > > scanimage for not giving the backend-provided stringlist options for the > > --mode option. > > scanimage prints the backend-privided values. Try e.g. > scanimage -d test --help > > As far as I can see, option mode is no special option in scanimage. > Ok, so the bug report is wrong.. I should know better than to trust user reports without testing their claims ;) > > So, my suggestion is that we also have VALUE strings defined in > > saneopts.h in a translatable fashion, to avoid the repetition of string > > constants, avoid the logical errors in translated versions, and avoid > > the inconsistancies that can develop between backends. > > That's a good idea. There is already a proposal for this in the SANE 2 > standard draft: > http://www.sane-project.org/sane2/0.08/doc014.html#s4.5.6 Excellent.. I suppose I should have checked that really.. > I'd just use these values and add them to saneopts.h (after the freeze). Will do. > Do we really need Color Lineart (or Lineart Color)? Better don't > mention that crazy mode anywhere :-) Just looking... the nec, sharp, and umax backends support it, and yeah they can't agree on which way around to put it.. That's a good enough reason to include it I guess :) Cheers, - Matthew Duggan
