https://bugs.freedesktop.org/show_bug.cgi?id=46355
--- Comment #5 from Lionel Elie Mamane <[email protected]> 2012-02-22 09:50:15 PST --- (In reply to comment #4) > Using 2 or 3 letter ISO codes looks scary to me. Do you think that an average > user knows the ISO code of his/her language? Not necessarily, but then (s)he is *still* better off than in the current situation. Let's take the example of a French/Korean bilingual person that uses a French language Windows. Thus the install UI is in French. French UI/helppack is checked by default, so the user wants to add Korean (his mother also uses this computer, and she doesn't speak French that well: recent immigration). Current situation: - Worst case: The user takes a cursory glance at list of languages. Looks sorted. Jumps to general area of where the searched language should be (all the languages that start with "c"), don't find it there. Sorry ma, no Korean support. Learn French or ask your brother to mail you a copy of ${COMPETITOR} that he buys for you in Korea. - Best case: *Maybe* the user notices they are not quite sorted, there are exceptions. So (s)he reads the whole list until finds "Coréen" among many other 'K*'. Checks "Coréen" and is on hir way. - Geek/developer case: *Maybe* the user notices they are not quite sorted. Had the same problem in a program (s)he develops so thinks "oh, they are probably sorted by ISO code or English name", let's look in these places. Finds "Coréen" quickly, checks it abnd is on hir way. Situation as I suggest: - Worst case: The user sees a bunch of cryptic codes and right next to them a localised name. Assumes they are sorted by code because the code comes at the beginning and that's how *everything* text is sorted: first according the the first letter, in case of equality according to next letter and then lather, rinse, repeat. Non-geeky user has *no* other notion of sort for text (strings). So resorts to reading the whole list until finds "Coréen" and is on hir way. - Best case: user knows (or "guesses") ISO code. Sees sorted by ISO code, jumps to general area, finds it quickly, confirms by the localised full language name guessed ISO code is correct and is on hir way. - Geek case: see "best case". Compare the "worst case" in the two scenarios. The worst case in my scenario is the best case in current situation. "Worst case" of current situation is avoided altogether. Besides, 2-letter ISO codes "often" match the country code of the country that gave its name to the language (if any). Users usually know that: it is the Internet country-code top-level domain. The 2-letter and 3-letter "T" variant has the advantage of being usually chosen to resemble the (Latin transliteration of the) language's name in itself: e.g. "fra" for French (Français), "deu" for German (Deutsch), sqi for Albanian (shqipe), cym for Welsh (Cymraeg) ... > I wish if it could be solved in the installer somehow, with a bit more clever > controls, that sort lists automatically. Yes, that would be the first choice. -- Configure bugmail: https://bugs.freedesktop.org/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are the assignee for the bug. _______________________________________________ Libreoffice-bugs mailing list [email protected] http://lists.freedesktop.org/mailman/listinfo/libreoffice-bugs
