Am 14.12.2016 um 17:31 schrieb Kevin Wolf: > Am 14.12.2016 um 17:17 hat Laurent Vivier geschrieben: >> Le 14/12/2016 à 17:08, Kevin Wolf a écrit : >>> Am 14.12.2016 um 16:58 hat Stefan Hajnoczi geschrieben: >>> IIRC, it's actually a known problem with not consistently using a UTF-8 >>> locale, and we avoid changing some locale categories because that would >>> break things like floating point representation in JSON (because printf >>> might end up using comma as the decimal separator). >> >> So the translation like the Chinese one is totally unusable? > > I just tried it: > > $ LANG=zh_CN x86_64-softmmu/qemu-system-x86_64 > > And yes, lots of question marks there... > > Kevin
It's good to know that there remains some work to be done after 2.8. :-) This is not a regression, so IMHO the status quo is good enough for now. Stefan
