Just food for thought ...
Just food for thought ...
It may put the translations slightly out of sync, but they will line up with time (and eventually we might be able to get the translation tools to flag changes to the source language). A brutal way of making sure they stay in sync is adding a nuke button to the EN_US version (You say the source language has been terrorized Mr. President? You can wipe out all translations, by pressing this button Sir. Then the spam engine will kick in and all embedded translators will be electronically informed to check translation interfaces for missing message keys so they can help rebuild their respective linguistic cultural heritages based on our ideals Sir).
I've come to realize that multilinguality requires more clarity in the phrases of the language you are translating from. I really empathize with the translators who are struggling to translate sloppy hacker speak that is not only cryptic but also out of context...
Actually the experience was good - I've been able to do a bit of internationalization work since those good old days!
where are your changes documented? We cannot expect 206 translators to do CVS diff checking against the message catalogs in order to keep track of their own terminological consistency.
It is great that the fact is now being recognized that "translating sloppy hacker speak" can significantly improve overall usability by discovering bad English original wording. But they must not be withdrwan just like a carpet from under the translators' feet.
(My two Eurocents in bad English).
I'm working in Russian and Japanese to make sure I get a fairly broad perspective grammatically before making recommendations. There will always be challenging phrases - a good example in Russian is that some number of nouns change their ending based on the LAST DIGIT of the preceding number, so 1 year, 21 year, 1001 year, nnn year-a, and
a new noun entirely (lyet) if the last digit is 5-9 (well 5 to zero technically). The trick in these cases is to invert word order to avoid such nuisances.
So in other words, what I hope to do is just keep good notes as I go through an entire language - maybe two, so when it's time to redo the english keys and create the common key structure in the acs-kernel, I'll actually know what I'm talking about. I'm hoping that other translators do the same so when it's time to caucus about how to fix the system we have a good dialog. I've been posting my basic thoughts (like ITEM is a vague term) in the forums because it's better than just letting it sit on a scrap of paper here.
In the meantime, I'm being careful but not fussy about what I translate. I think every one of us who takes on a language will have to go back and make all the phrasing consistent after a first pass anyway.