Corrupt national characters again

All announcements, questions and issues related to the TsiLang Components Suite.
Post Reply
Xyzzy
Posts: 13
Joined: Tue Jan 06, 2009 11:52 am

Corrupt national characters again

Post by Xyzzy »

Hello

I've got a new version to translate for a software, and a lot of Polish national characters are again corrupted! I send OK file for version 6, and version 7 has these corrupt characters :/
I can send you files to compare.
Of course if the problem is with handling files on the software's author part, I will pass any instructions to him.
Xyzzy
Posts: 13
Joined: Tue Jan 06, 2009 11:52 am

Post by Xyzzy »

The same story repeats. Collectorz.com Game Collector v4. In version 3 file I sent characters are OK. In version 4 released there are lots of garbage: 'ł' changed to '3', 'ś' changed to 'o', 'ż' changed to '?', 'ć' changed to 'a' (that's my favourite, 'a' being the most frequent letter in Polish) and so on... :evil:
Not all occurrences, but enough for a f****** lot of extra work.
isiticov
Site Admin
Posts: 2383
Joined: Thu Nov 21, 2002 3:17 pm

Post by isiticov »

Hello,

Thank you for your message. We can't know what the author of software does that leads to national characters corruption. May be there is used a font name for Polish language which isn't supported on their PC and this forces the characters to corrupt or so on. May be the charset setting for Polish language is set to DEFAULT_CHARSET and on your PC it maps to EASTEUROPE_CHARSET but on the authors PC to another one.
I would recommend you to contact author and check this with him.
Best regards,
Igor Siticov.
Xyzzy
Posts: 13
Joined: Tue Jan 06, 2009 11:52 am

Post by Xyzzy »

Well, if there is possibility that I send translation with character code X and it is changed by your software to character code Y (in .sib file, I don't mean display), that is bug in your software. So simple.
And it doesn't matter if it is result of 'real' bug or result of workflow that allows such corruption.
isiticov
Site Admin
Posts: 2383
Joined: Thu Nov 21, 2002 3:17 pm

Post by isiticov »

Hello,

Thank you for your message and your feedback, we really appreciate it.
If software author used non-Unicode controls in his application and created non-Unicode SIL\SIB files, it is quite complicated task to try to keep all national characters under all locales as there are a lot of very different locales. When anybody opens a SIL\SIB file with ANSI translations on their locale they can easily edit the translations that correspond to their locale but at the same time this might corrupt the national characters from another locale if there are incorrect or insufficient data stored in SIL\SIB file. SIL Editor is Unicode application and it edits translations in Unicode mode trying to convert the ANSI data (if used) to Unicode and vice versa. When there is any incorrect combination of settings used this may lead to corruption.
For example our tool - SIL Editor is translated in several languages uncluding German, Russian and Hungarian. All the translations stored in external file, which you can also translate to Polish (as well as many other users over the world who translated it to their language). And the translations never been corrupted for years because of proper settings and information stored in the file.
Hope this helps.
Best regards,
Igor Siticov.
Post Reply