Re: [Gnucash-changes] Fixing UTF-8 support in X sessions

2005-10-17 Thread Didier Vidal
Also, instead of caling bindtextdomain, we should call bind_textdomain_codeset Didier. Le lun 17/10/2005 à 19:41, Derek Atkins a écrit : > A better solution would be to change the gnucash code to use > the g_locale_to_utf8() API in the gnucash locale settings. > We should /not/ force gnucash into

Re: [Gnucash-changes] Fixing UTF-8 support in X sessions

2005-10-17 Thread Christian Stimming
Am Montag, 17. Oktober 2005 21:05 schrieb Didier Vidal: > Actually, adding UTF-8 to an arbitrary locale might be dangerous. It > should only be a temporary workaround, but should not make its way to a > release. > For instance, on fedora2, the Japanese locale doesn't have an utf8 > counterpart. (By

Re: [Gnucash-changes] Fixing UTF-8 support in X sessions

2005-10-17 Thread Derek Atkins
Quoting Christian Stimming <[EMAIL PROTECTED]>: prompt> locale -a | grep jp ja_JP.eucjp Yes, thanks for pointing this out, and this is certainly true for any of the gnucash series. So the original patch cannot go into any release. Well, I backed out the g2 change already. I agree, it's not

Re: [Gnucash-changes] Fixing UTF-8 support in X sessions

2005-10-17 Thread Didier Vidal
Actually, adding UTF-8 to an arbitrary locale might be dangerous. It should only be a temporary workaround, but should not make its way to a release. For instance, on fedora2, the Japanese locale doesn't have an utf8 counterpart. (By the way, Japanese translations of gnucash exist... so we must hav

Re: [Gnucash-changes] Fixing UTF-8 support in X sessions

2005-10-17 Thread Derek Atkins
A better solution would be to change the gnucash code to use the g_locale_to_utf8() API in the gnucash locale settings. We should /not/ force gnucash into a utf8 locale. -derek Christian Stimming <[EMAIL PROTECTED]> writes: > Oops, Neil, > > sorry for that -- my first complaint obviously was wro

Re: [Gnucash-changes] Fixing UTF-8 support in X sessions

2005-10-17 Thread Christian Stimming
Oops, Neil, sorry for that -- my first complaint obviously was wrong. Turns out I didn't read the generate-gnc-script in full so that I didn't realize the commands are part of the here-document. So the "dependence on LANG at compile time" doesn't exist, and that is obviously a good thing. Ne

Re: [Gnucash-changes] Fixing UTF-8 support in X sessions

2005-10-17 Thread Christian Stimming
Dear Neil, that change is totally b0rken. Please revert it. Firstly, your change means that the generated script depends on the LANG setting at compile-time. This is fundamentally wrong. We cannot assume any relation whatsoever between the LANG at compile time and LANG at run time. So even if