Le dimanche 09 septembre 2007 à 15:35 +0200, Julien BLACHE a écrit : > Petter Reinholdtsen <[EMAIL PROTECTED]> wrote: > > Hi, > > > I guess I learned something new, then. I thought the .mo files was > > mmapped into the programs and looked up using fast and arch-specific > > methos. If this is not true, I can change my packaging practice. :) > > I don't know the specifics on gettext's implementation, but this file > from the gettext manual offers some hindsight on the file format: > <http://www.gnu.org/software/gettext/manual/html_node/MO-Files.html> > > I don't see anything arch-specific at first glance, and they discuss a > couple of issues like hash tables and alignment mostly to end up > telling that it doesn't buy much in the end :)
That's also my opinion, but I'm going to ask a gettext expert. For now I'm lowering the bug severity. Thank you both for your input :)
