Hi! I just noticed that python has native gettext support (since 2.4), it would be nice towards translators to switch the custom i18n support into gettext. So that translators can work with the common po files and all the accompanying po-related toolchain, to be able to easily handle updates to the origin strings and similar, for example the Spanish translation seems out of date. Thanks, Guillem
2013/9/4 Guillem Jover <guillem@debian.org> That would be indeed good. I wonder why you think that the Spanish translation is out of date. Even though I'm not a translator myself, I took care to make the full translation into Spanish and tried to keep it up to date. As far as I know it is. Did I miss something? What about the translation seems out of date? Greetings and thanks, Miry
Hi! Sorry for the alarm, you are absolutely right (I just checked what I thought I had seen before), I think I might have gotten confused while comparing the contents from en_us.py and es.py from two editor windows. Thanks, Guillem
2013/9/4 Guillem Jover <guillem@debian.org> Thanks a lot for your efforts and suggestions, btw :) Greetings, Miry
Hi! I was initially looking into using the .po files but somehow I got distracted by the complexity of the task and I went for using python dictionaries, lists and functions instead. I am afraid I won't be able to change it very soon, but I will try to learn the library and see how easy it would be to replace the current translation files with the new solution. Kind Regards, Ireneusz Imiolek
Hi again! Just a quick update. I have completed the transition to using gettext for i18n. This is kind of non standard implementation - but I have chosen the quickest way to convert the application with what I already had. Most of the 'messages' are now included in .po files. I have still kept some stuff in the custom language files, such us numbers (need them from zero to one hundred - so it's easier to generate them automatically) and other stuff that was hard to implement with gettext. The new version also solves bugs: #721829, #721830, #721899. @Guillem: You mentioned under one of your bug reports ( http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=721829) that you had started working on the Catalan language. Did you manage to complete any of those translations? If you have could you please send me what you have so I can convert it to .po file - this would save you some time copy/pasting it into new file. That's just if you still have time to complete this. If you would like to check how it works and download the new version with .po files - the pre-release source code is available at: https://drive.google.com/folderview?id=0BzzRSdFJhjQ8NHBhYTZOTGpSZE0&usp=sharing @Miriam: I mentioned I'd send you the .po file, but since quite a lot has changed because of that transition it would be easier if you get the whole package so you can test your translation. When you finish editing the .po file and save it with Poedit, please move the generated .mo file to mo directory to test your changes. Thanks, Ireneusz Imiolek
Hey! Great! Sure. I think some other of the messages in the <LANG>.py files would also be good to have in the .po files, at least the hours and time. The color ones would benefit from gettext context functions, but they do not appear to be present in the python interface, unfortunately. Also the po file names would also benefit from being made more generic, so that the fallback code can choose a closer language when available. For example using es_ES.po means that a Mexican using es_MX.UTF-8 as locale would not fallback to use the Spanish language (es.po) version, which would be better than no translation. The same for a Brazilian (pt_BR) or for a non-US or British English speaker like a Canadian (en_CA). So I'd recommend renaming es_ES.po to es.po, pt_PT.po to pt.po and en_GB.po to en.po (and leave the other one untouched as a specialized locale). (Should probably move this into a new bug report, but anyway.) I didn't finish yet, I paused when I got stuck with the flash cards. I've attached the partial stuff that I got in patch form. I'm missing images for the three words with J, K and U letters. Also for Ç, had to use a word with the letter inbetween (Calçat = Shoes), there's no meaningful words in Catalan starting with that letter, hope that's fine. Having the flashcard images in separate image files would make it easier to add new ones or to assign the correct name to the correct image, but I'm not sure if there's a particular reason to have them as sprites. For J (Jaguar), I found this one: <http://openclipart.org/detail/167328/jaguar-by-frankes> but then I think it might be a bit difficult word, probably would be better Joguina/Joguets = Toy/Toys, but finding a nice, easily recognizable image for toy seems a bit difficult. Jardí = Garden could be another option, but finding an image for that might be difficult too. For K (Kiwi), I found this one which seems to match the style of the other flashcards (I realized later on there's already f_img6a.png): <http://openclipart.org/detail/129397/kiwi-by-gnokii> For U (Ull = Eye), I found this image that I think is nice: <http://openclipart.org/detail/3469/-by--3469> That looks nice! I noticed the Spanish flashcard for violin is wrong. And I'll try to finish the Catalan translation when I've some time. Thanks, Guillem
Hi again, I think some other of the messages in the <LANG>.py files would also be good to have in the .po files, at least the hours and time. The times would be fine, but I chose to leave them out because of the way the game works - the program relies on the fact that only some entries are translated depending on the way time is told in a particular language and if I include them all in the po files, all of them will get translated - this will make the choosing of the correct version confusing. (ie. should the 5:35 be read as 35 past 5 or 25 to 6 or maybe "five thirty three". This part is usually done by me based on info from translators). I guess I need to update the translation instructions to avoid confusion. In this case I am considering replacing them with 18 lines of translatable text, ie: 1. "add more blue" 2. "too much blue" 3. "blue is OK" for each of the 6 colours. In fact it might be a better option. Also the po file names would also benefit from being made more Well, the fallback is already built in with the code. First I check if there is a full locale available, if not - check for first 2 letters, ie, if es_MX.po is not available it will look for es_*.po and use the first matched language (es_ES.po in this case) - if no language matches the criteria for fallback then the en_GB.po is used. Well, thanks for that one - I was doing all flashcards myself for all languages so I wasn't expecting anyone to go so far with this - well, I wouldn't expect anyone to go searching for images either :) . With regards to the Ç - in some languages I used middle letters as well - ie. in Polish no word starts with ą or ę, etc. Having With sprites It takes a while to load the game initially but then going between letters is much quicker, whilst if I was to load them when clicked it would take more time. I thought delaying game on start is better than after each click. I will separate them and see how it works if it takes too long I will just add the images you have mentioned to the sprite. Thank you and sorry I have overlooked it. Wrong reference to image position - and somehow I have missed it while testing. As soon as I get all these issues solved I'll update the temporary folder with new "pre-release" version. Thanks again for your time. Ireneusz Imiolek
Hi, Just a quick update... bug fixes: - color related stuff reworded and added to po files - flashcards updated, actually I did drop use of sprites for this particular activity - as you mentioned - it's much easier to add images - they still need some compression treatment because as of now the size of the package increased from 3.7MB to 5.1MB. This change doesn't seem to affect the performance - so it will stay. I have added your suggested images - very nice finds - Thank you. I used the little toy wagon for the Joguina - hope it makes sense, otherwise will probably have to use Jaguar. - the Violin error in Spanish fixed - the submitted patch file adapted to new version - the ca.py 'converted' to ca.po - the English word list is used for now, I will get this updated soon. - new test version updated on google drive: https://drive.google.com/folderview?id=0BzzRSdFJhjQ8NHBhYTZOTGpSZE0&usp=sharing Thanks, Ireneusz Imiolek
Hi!! Sorry for not having been responsive for the last couple of months. I'm slowly catching up with the pending tasks in Debian. Just wanted to drop a line to say sorry and tell you that I'm gonna try to work on it as soon as I can. Greetings and Happy Xmas, Miry 2013/11/11 Ireneusz Imiolek <imiolek.i@googlemail.com>: