Minimum translation percentage for Plasma 5 release

[Albert Astals Cid]

El Dimarts, 12 d'agost de 2014, a les 11:56:40, Māris Nartišs va escriure:
> Hello Vincenzo,
> I read Martin's reply as "it would not hurt to have partial translation".
> 
> Keep in mind the motivating factor. I joined KDE, QGIS and GRASS GIS
> translation efforts because the shipped translations were incomplete,
> imperfect or even bogus. Artificial percentage doesn't provide the quality
> of translation. 

Right, percentage is a "weak" metric for quality, translation can still be 
totally broken at 100%

> Those, who are following this mailing list, probably have
> noticed "languageo FOO is FUBARED and I want to take over it to fix it"
> type e-mails. Thus if the choice is between "partial translation" vs "no
> translation", I would vote for "partial translation".
> 
> Although I liked  Franklin's idea - 70% = warning barrier and like 50% =
> hard barrier. If a language fails below 70% - warn the maintainer. If there
> is no response from the maintainer within N release(s), stop shipping as it
> is unmaintained. Thus there would be no 69.9% issue and maintainers
> struggling to keep up the % would still have a chance to get language
> shipped as long as they care.

Thing is as I mentioned, we should not punish the user because a maintainer 
stopped caring, at 50% a software may still be very much useful.

> A "no ship" barrier still would be nice, as there are some languages having
> 0% translated strings. Listing such language as an "one of KDE (SC)
> languages" would make us a laughing stock.

Sure, I agree that 0% makes no sense, so let's say something as low as maybe 
15% to make sure it starts being useful, i think with 15% you can have the 
most used buttons, etc translated already (of course you can also have obscure 
stuf translated, as said percentage is a weak metric)

> Sorry, Albert, this offer adds some extra work for you.

Why would it be extra work, we already have this concept ;) And i'm not doing 
the plasma 5 releases anyway :D

Cheers,
  Albert

> 
> Just my 0.02,
> Maris.
> 
> 2014-08-12 11:30 GMT+03:00 Vincenzo Reale <smart2128 at baslug.org>:
> > In data martedì 12 agosto 2014 10:17:59, Martin Schlander ha scritto:
> > > Mandag den 11. august 2014 20:37:28 skrev Albert Astals Cid:
> > > > So I am leaning towards "no minimum", but I certainly welcome all
> > 
> > comments
> > 
> > > > and am open to change.
> > > > 
> > > > Opinions?
> > > 
> > > Hmmm, as a translator it's probably in my interest not to have any
> > 
> > minimum,
> > 
> > > even though I don't plan to be flirting with 70% completion very much.
> > > 
> > > The question is whether very incomplete translations might hurt
> > 
> > KDE/Plasma
> > 
> > > in the eyes of users, more than having no translations at all. But I'm
> > > leaning towards no, probably not.
> > 
> > Hi all,
> > I fully agree with Martin's statement.
> > Incomplete translations are really ugly to see. It's better having no
> > translations.
> > 
> > Regards,
> > Vincenzo