Hi Jordi, On Thursday, 2006-10-26 14:23:41 +0200, Jordi Mallach wrote:
> I've re-read your previous message, and I realise I didn't understand > what you meant with: > > > | | My concern is that can be used as a some kind of non-scientific > | | justification for some ultra right wing Spanish nationalists who > | | claim that "Valencian" and "Catalan" are two different language > | > | Unimportant, as long as neither ISO nor IANA have a language code > | assigned, it's no distinct language for all standards compliant > | applications. > > I thought you were talking about the "language war", not about > supporting the Catalan-related variant in OpenOffice.org. No, from a developer's technical point of view I'm not concerned with "language wars" or any separatist's or political movement.. my concern in this regard is only about complying to standards OOo uses, ODF and underlying RFCs, ISO, whatever. > I see I'm alone in the desert right now. I'll have to do an internal > build for a weird "valencia_ES" locale though, so I can still use some > advice from the list. If this is only about UI localization and not document content, we could assign a LangID and create a mapping to a private "x-catval" or similar, not listing the language in the language listbox. Note that in a multi-language installation it would not be listed under Tools.Options.LanguageSettings.Languages "Language of User interface", or shown as "Unknown" at its best. However, there are known issues with a legacy parser and the xml:lang attribute used in the menus' .xml files, the parser refuses anything other than an ISO 639-1 alpha-2 code for the languag part, this is already a problem with ISO 639-2 alpha-3 codes, let alone RFC 3066 x-... private :-( See also http://qa.openoffice.org/issues/show_bug.cgi?id=70181 > > > // { LANGUAGE_CATALAN_VALENCIAN, "ca", "ES" }, // Modifier > > > names not supported yet ([EMAIL PROTECTED]) > > > > This would introduce ambiguousness to an already existing mapping: > > > > { LANGUAGE_CATALAN, "ca", "ES" }, // Spain > > (default) > > > > and is not "round-trip-safe", conversions back and forth between that > > mapping would not produce identical identifiers, which is essential for > > using it in attribution and storage. > > Yeah, that's why I commented it. Technically, if a document edited with > a hypothetical Valencian OOo, I wouldn't mind if the document had the > same language tag as the purely Catalan version. The only interesting > changes are UI wise: dictionaries, output files and such could be > exactly the same. Ah, I thought these were also subject to differentiation. Then the x-catval could be the way to go. UI localization and document content are independent. There may be some places where this isn't the case though and some default is obtained from the UI locale instead of the working locale, in this case it would probably fall back to en-US. > > > { LANGUAGE_CATALAN_VALENCIAN, "ca", "valencia" }, // > > This would not produce a valid language/country pair > I see. I tried mimicing the Serbian Latn/Cyrl examples in the file. These are nasty kludges, workarounds, though based on former, now outdated, ISO codes. I wouldn't introduce more of such kludges. > The Valencian government produces an educational GNU/Linux distribution, > LliureX, which for the usual political reasons did not use a standard > locale like the proposed "[EMAIL PROTECTED]". Instead, it went the non-ISO > 639 route of inventing a "valencia_ES" locale, which doesn't stomp on > the ISO 639-2 namespace, and is still usable by Gettext. But may clash with some standards aware applications. > If there's absolutely no way I can integrate such locale in > OpenOffice.org proper, then it won't be possible to ever have a public > Valencian build for Softcatalà (Jesús' and my translation organisation), > which sucks. If there is a way, I'm interested in knowing what it'd > imply. > > If it didn't work out, is it possible to reserve an unofficial LangID > for us even if it's never used officially? That way, I could work on > custom patched builds using "my" langid knowing I'll never stomp on > anyone's feet in the case a new locale appears/gets assigned. As mentioned, the LangID isn't the problem. I can assign one, and a mapping to some RFC 3066 code. The problems start with the xml:lang attribute parser during build time.. > > > Is this going to work at all? > > > > Not yet. Vaiants would need modifications to the file format and several > > modifications to the source code, probably throughout the entire > > application suite. > > Yes, Ivo told me about a year ago, and warned me that this would > probably wouldn't happen in some time. Can you estimate how much? _At least_ one more year. > > Sorry I can't tell you "hey, that's easy, just do this and that" ... > > I hope I've explained the situation a bit better now. No matter what, > I'll have to do a custom, hackish if needed, build for LliureX, using > whatever locale I need to use. The idea was to have compliant support in > the official repository though, so that a "good" Valencian version can > be produced. Note that if your private solution additionally to UI localization is also used in document content, when interchanging documents with other OOo releases the document can't be properly processed, and saving it again will discard locale information. Eike -- OOo/SO Calc core developer. Number formatter stricken i18n transpositionizer. OpenOffice.org Engineering at Sun: http://blogs.sun.com/GullFOSS Please don't send personal mail to this [EMAIL PROTECTED] account, which I use for mailing lists only and don't read from outside Sun, use the [EMAIL PROTECTED] account instead if really needed. Thanks. --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
