Talk:Localization/Concepts/Transcript

From KDE TechBase
Revision as of 21:33, 5 January 2009 by Ilic (talk | contribs)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)

I must say that I prefer this approach to the l20n approach. But one concern. Can we try to avoid the msgstr approach. Yes I know it works but it took a long time to undo _n: and _: from KDE3 and it would be nicer if we didn't need to do that. I'd suggest asking the PO developers to actually allow this feature in an extension to the format.

I realise that that might cause problems with editors since they would need to support this, but in the long run I think its a better approach.


I too think having a new field in PO format (e.g. msgscript) would be better approach, and in personal communication Gettext's maintainer, Bruno Haible, has already shown some interest. However, there would be many more details to work out to bring scripting into the Gettext itself, as Gettext and PO format aim for generality across languages and platforms, as opposed to singular environments like KDE.

On the other hand, in my opinion having extra stuff embedded into msgstr is much, much less "evil" than embedding into msgid. For one, an unwary translator who knows nothing of the scripting, also can't do any damage (while when context was embedded in msgid, people sometimes translated it as if it were part of the message).

Ilic 21:33, 5 January 2009 (UTC)