support for documentation translation. In short, it is _conceptually_ easier to fully localise an application (GUI/Docs in various formats) by using an exclusively .xliff based process than by using an exclusively .po based one.I think this is a false statement. There are tools to handle documentation as po files, and I believe kde is using this successfully for all theirdocumentation, and has done so for some years.
I have no doubt there are tools to handle documentation as po files. I am just saying that gettext was meant to deal with application produced messages before specific formats were developped to address issues related to localisation.
As far as translation management is concerned though, it seems to me that translation variants are better handled by xliff and this specific item should greatly enhance the translation process in Debian if properly implemented:Why do you think that? I believe po files has good support for management.And this has been one of the strong points vs. other formats, viz. java-bundles. Where is xliff strong in management? Having more than one version in one file could be seen as confusing.
And sorry if I confused you. In my earlier mail I argued about multiple tu variants, but as I also wrote, xliff is a bilingual format thus containing only 2 variants: target and source. The multiple variants are included in tmx.
But whatever, xliff includes process related metadata which is one of its strong point. The standard is online as well as plenty of articles dealing with l10n and xliff. Feel free to check them.
I don't know much on xliff, but the gettext system has standard programsfor translation memory and compendia - which seems to be similar to the functionality you claim that only xliff has. Could you explain the functionality of each of these components more?
I am not claiming "only" xliff has any functionality in the part you replied too, only that there are a few sets of standards that have been developped with integration in an xml based l10n process of which xliff is a part. SRX is a segmentation standard, TBX a glossary standard, TMX a translation memory exchange standard.
Each standard is clearly specified on the web with plenty of related information that you can read if you are interested.
One question: can you have other character sets than Unicode in xliff?
Why don't you check yourself ? Also, this mail may mean something to people on the po side:http://mail.gnome.org/archives/gtranslator-list/2005-January/ msg00001.html
JC