Lynn Allan wrote:

Sword-devel'ers:

There are quasi-internationalized version of the LcdBible software available
at:

Spanish RV - NT:
http://prdownloads.sourceforge.net/lcdbible/LcdBibleSetup_sparv.exe

French LSG - NT
http://prdownloads.sourceforge.net/lcdbible/LcdBibleSetup_frelsg.exe

German Luthern - NT:
http://prdownloads.sourceforge.net/lcdbible/LcdBibleSetup_gerlut.exe

Some notes and caveats:
* The Bible book names were borrowed from The SWORD Project for Windows
using the options-translation
* For now, an end-user could only have one trans at a time on their computer
* I just used google-trans for guessing at the equivalent of "Search For" or
"Search Phrase". I ignored, for now, the issue of word order.
* Once you get the hang of it, it's is nearly trivial to do this quick/dirty
internationalization ... an hour or so per trans (in queue: Swahili, GerBen,
KLV (Klingon?), Czech, Turkish, Uma, Bulgarian, Greek)

How are you actually doing the i18n? From this end it sounds like you might actually be maintaining separate codebases or something rather than using different resource files to swap strings. Maybe I just missed something in an earlier post.


If you're using Sword, there are some i18n routines built in I believe. Aside from that, if i18n/l10n are real goals, you should consider using ICU, since its features are much more robust than those in Sword. You'll definitely want to consider ICU for InVerse, since it has some very nice international calender data (localized weekdays & months, different religious calendars, local holidays, etc.).

* I only wrestled with uncompressed Bible texts
* The installation routine words aren't translated.
* The titlebar isn't translated
* I'm not clear if all these use the same versification, but assumed this
was the case

If they're Sword modules, they do, but in reality they don't. (French versification being famously distinct from English.)


* All chars are 8-bit ascii rather than unicode or multi-byte

As Hugo pointed out, ASCII is 7-bit. On Windows, codepages will provide you much pain. You're actually using Windows Codepage 1252. All of the languages you mention use this codepage. Of the translations you list as being "in queue", 4 use CP1252, the other 4 use 4 different codepages. This is something you'll need to address very early on if you intend i18n to be a core feature. I strongly recommend switching to Unicode and specifically using UTF-16 as your encoding, since WinNT uses it natively.


* LcdBible is only for Window with Win98-Second Edition or later, not Linux

Feedback appreciated. I'd also like to check if it would be appropriate to
request feedback on the sword-support list.

--Chris



_______________________________________________ sword-devel mailing list [EMAIL PROTECTED] http://www.crosswire.org/mailman/listinfo/sword-devel

Reply via email to