On 19.10.2012 15:50, Zeeshan Ali (Khattak) wrote: > On Fri, Oct 19, 2012 at 4:13 PM, Michal Privoznik <mprivozn@xxxxxxxxxx> wrote: >> On 19.10.2012 15:07, Zeeshan Ali (Khattak) wrote: >>> On Fri, Oct 19, 2012 at 12:56 PM, Michal Privoznik <mprivozn@xxxxxxxxxx> wrote: >>>> On 19.10.2012 05:49, Zeeshan Ali (Khattak) wrote: >>>>> diff --git a/po/LINGUAS b/po/LINGUAS >>>>> new file mode 100644 >>>>> index 0000000..af407ee >>>>> --- /dev/null >>>>> +++ b/po/LINGUAS >>>>> @@ -0,0 +1 @@ >>>>> +# Keep this list sorted alphabetically >>>>> diff --git a/po/POTFILES.in b/po/POTFILES.in >>>>> new file mode 100644 >>>>> index 0000000..88dfb7e >>>>> --- /dev/null >>>>> +++ b/po/POTFILES.in >>>>> @@ -0,0 +1,29 @@ >>>>> +osinfo/osinfo_db.c >>>>> +osinfo/osinfo_deployment.c >>>>> +osinfo/osinfo_deploymentlist.c >>>>> +osinfo/osinfo_device.c >>>>> +osinfo/osinfo_devicelink.c >>>>> +osinfo/osinfo_devicelinkfilter.c >>>>> +osinfo/osinfo_devicelinklist.c >>>>> +osinfo/osinfo_devicelist.c >>>>> +osinfo/osinfo_entity.c >>>>> +osinfo/osinfo_filter.c >>>>> +osinfo/osinfo_install_config.c >>>>> +osinfo/osinfo_install_config_param.c >>>>> +osinfo/osinfo_install_script.c >>>>> +osinfo/osinfo_install_scriptlist.c >>>>> +osinfo/osinfo_list.c >>>>> +osinfo/osinfo_loader.c >>>>> +osinfo/osinfo_media.c >>>>> +osinfo/osinfo_medialist.c >>>>> +osinfo/osinfo_os.c >>>>> +osinfo/osinfo_oslist.c >>>>> +osinfo/osinfo_platform.c >>>>> +osinfo/osinfo_platformlist.c >>>>> +osinfo/osinfo_product.c >>>>> +osinfo/osinfo_productfilter.c >>>>> +osinfo/osinfo_productlist.c >>>>> +osinfo/osinfo_resources.c >>>>> +osinfo/osinfo_resourceslist.c >>>>> +osinfo/osinfo_tree.c >>>>> +osinfo/osinfo_treelist.c >>>> >>>> Since there is nothing to translate yet, we should drop this addition. >>> >>> Christophe pointed out the same in his review. The thing is that there >>> is no translation yet. When a string is marked for translation or a >>> new translated string is added, its very likely that developer will >>> forget to add the enclosing file into this list. So IMHO its better to >>> add them all already here since it'll only add a bit of build-time >>> overhead. >> >> 'make syntax-check' (which should be ran prior every git-push; and I ran >> it even prior posting a patch) should have failed in this case. And it >> indeed did. > > You mean if I mark a string for translation and don't add the file to > this list, `make syntaxcheck` will fail? Yeah, that's exactly what I meant. > >>> AFAIK, thats how most projects do it. >>> >>> Talking of this, I was going to ask: Should we or should we not be >>> translating error messages? Some UIs might show it to the user and >>> ideally every message shown to user must be in his/her locale. >>> >>> >> >> Error messages should be translated. Or better to say - there should be >> an opportunity to translate them. That's how we do it in libvirt. >> However, not the debug messages - they are always in English as most >> users don't see it - unless they are developers playing with it. > > Yeah, I wouldn't translate debug messages either. I'll mark error > messages for translation then. BTW, how do you get people to translate > libvirt? Boxes gets translators for free, being part of GNOME and > hosted on git.gnome. > Mainly from Fedora. But I am not so familiar with the process to get any deeper. Michal
