[info] Translation extract from the old website
I would like to inform you that I have written a script to extract the translations from the old website that converts (cleans) the strings correctly. I have managed to make 20% progress on 16 untranslated languages. It may not seem like much, and the page looks half translated. But the important thing here is the morale and the simplicity.
Languages are: BR, CA, CS, GL, HU, ID, IT, JB, KR, KW, NO, NS, PT, RO, SZ, UK
Now to explain why: As a translator, I know what it is like to have to go through a complicated process to start translating. Also, when I saw that the translations from the old site were "lost" when you switched to the new one, I didn't feel like translating again. Yes, it is a better site! And the Weblate integration was a very good decision. But when you try to continue from there, the language is there, but you can't do it as such, because it's empty at the beginning. I knew that there was old data and that it could theoretically be extracted and put into the new format. So I've been waiting for someone to do it for everyone. I waited for months, years, two years.... It didn't happen! Then one day I took the trouble to start translating the page and copy the old text by hand. Luckily the longer ones didn't change.
Recently, when I was testing the site on my old server, I came across the old one for testing. I was surprised at how many languages there were and started to investigate: How many old translations were used on the new site? The result: Besides SL Slovenian, DE German and RU Russian (which I copied all by myself), I only found NN Nynorsk. I may have missed some, but most of the translations were done from scratch, and many people probably didn't even know that old translations existed. I found it annoying to have to learn a format I'd never seen (PO) and I'm not a layman (I have an IT background and SVG was quite familiar). I'm not saying it's bad or difficult, no! It's very useful and allows the complicated plurals of Slavic languages to be implemented correctly. I was impressed!
But for non-technical people who want to contribute, it seems too complicated. Weblate has nothing to translate (yet), figuring out what to do in GIT takes time (and frustration) that many don't have (not counting recent GIT blocks). I have tried to do everything necessary to get started twice in the last 2 years and failed (to read the friendly manual - hoping for an easy way).
If someone had done 'all this' before, I'd be so happy that I'd start translating as soon as I heard about it. Now I am the one who has done it for others. Perhaps this is an offer of help to the old me, or rather to others in the same situation. It took me a whole day to create custom filters in Python, but imagine how much work and nerves this will save: It was * worth it!< Now they can just register to Weblate and recieve additional informations (example pictures, comments, other languages examples) meanwhile they are translating.
Now they can simply register with Weblate and receive additional informations (sample images, comments, examples in other languages) while they translate.
I'll prepare the merge request when I have the XX-credits.json with the old translation names ready and check a few last things (it's already working on the test page).
PS: Should we make an announcement to notify the old translators when it's merged?