Follow-up Comment #4, bug #11279 (project wesnoth):

> But it is sheer nonsense to try to make a pot-update of a 
> release tarball. Part of the creation of the tarball is 
> automatically running make update-po, so the files in there are
> as recent as possible for the very release.

Very strange again. I noticed untranslated strings in the campaign "The
Hammer of Thursagen", e.g. "Dwarvish Masked Steelclad". I looked at the
source and saw this string in the file
data/campaigns/The_Hammer_of_Thursagan/units/Dwarvish_Masked_Steelclad.cfg.
The string even had the gettext prefix but did not show up in any of the .po
files. Therefore it did not seem to me to be sheer nonsense to run
update-po.

> And the version of gettext used does not change too much, I was
> basically using every version from 0.13 onwards already, there
> was never such a problem.

I am doing i18n & l10n for many years, for many projects and different
versions of gettext or intltool have been the culprit several times. Last
time was when I was working on http://www.gcompris.net/. Ask the author
(Bruno Coudoin) for details.

> Even if make update-po (as a general pot-update) would not work
> for the tarball this is irrelevant since the tarball itself is
> never used for this stuff, only svn checkouts have to work with
> it.

If a source tarball just contains the mininal set of files for compiling the
binary or contains everything so that working with the _source_ is actually
possible is a good base for a flame war. You decided to go with the first
option. Fine.
At least I can confirm that the problem does not exists when using the svn
checkout.

    _______________________________________________________

Reply to this item at:

  <http://gna.org/bugs/?11279>

_______________________________________________
  Nachricht geschickt von/durch Gna!
  http://gna.org/


_______________________________________________
Wesnoth-bugs mailing list
[email protected]
https://mail.gna.org/listinfo/wesnoth-bugs

Reply via email to