Translating with Quality
Recently I get involved again in Arabeyes in the KDE Translation project. So far I didn’t do much more than syncing files between the Arabeyes CVS and KDE CVS. I tried to open some files, and look into the translation in it. To be frank I was very unhappy. It is not KDE only, but as well other translation projects. And part of what I was unhappy about is strings I translated 2 years ago. Now, I know that the translators are doing their best, but the whole process and Quality checking and assurance is not working. So, all the time and effort spent into translating programs and user interfaces in Linux may be a waste of everyones time without proper QA.
I learned recently about a QA Committee in Arabeyes, I am looked for pages about that QAC in Arabeyes pages, and I didn’t find much, accept for 3 pages in the Wiki with the Agenda and minutes of the QAC meetings. I was talking to a friend of mine who is working with a company that deliver a Linux Desktop solution for the enterprise, he noted that he is spending more time is checking the quality, and fixing the errors in the translation, than it would need him to rewrite/translate the whole thing. I Agree with him, and the emphasis should be on quality of translation, not on “how much is done”.
I am unaware of any Linux user that uses the Arabic-translated user interface in either KDE or Gnome. Although they had been present for around 2 years. Is the reason the quality of the translated interface ? or the people I know prefer english over Arabic for a user interface.. I am not sure
It would probably be better if those “corrections” were given back to the translation team so he wouldn’t have to keep making them on each release.