Re: proposal for splitting translation tests into a separate project
David Shea <[email protected]>
| Newsgroups | gmane.linux.redhat.anaconda.devel |
|---|---|
| Message-ID | <[email protected]> |
On 11/19/2015 03:06 PM, Chris Lumens wrote: >> One more for this list: give the boot to languages whose percentage of >> translated strings falls below a certain threshold. For example, anaconda >> includes Iloko, which is 2.3% translated in f23-branch. This is not useful. > How do we kick something out? Just remove it from po/LINGUAS? Yes. I think the last language we did that to was Hebrew, since the guy translating Hebrew let us know he was quitting. > >> I don't know whether we should bother looking for the opposite, when someone >> suddenly translates something that was previously untranslated. You would >> think someone give us a heads up, but I can only think of that ever >> happening once in the past. > We're never going to hear if something passes a threshold or not. At > the most, we might find out if a whole new language needs to be added, > but that's pretty rare. I was thinking of a case of, for example, a translator going through anaconda strings in Zanata and seeing a bunch of untranslated Hebrew strings and saying "hey I know Hebrew I can fix that" but then not telling us. We can test for that, I think. I can't remember whether transifex did or not, but zanata pulls down everything regardless of LINGUAS, so we could look for languages not in LINGUAS that are above a high threshold of percentage translated. But yes, I agree that this would be rare.