Hi Jean-Charles and all,
On submitting latest PO to the TP robot, it was at first rejected
because 2.15.29 in the project-id line, which doesn't match the file
name for 126.96.36.199
My surprise is to see that latest accepted file for plain 2.15.37 has
the same line "2.15.29". How was it accepted?
Anyway some a llittle bit wrong is here if we are generating those
with .29. What do you think?
Francisco Vila. Badajoz (Spain)
www.paconet.org , www.csmbadajoz.com
Le 29/04/2012 10:56, Francisco Vila disait :
> Hi Jean-Charles and all,
> On submitting latest PO to the TP robot, it was at first rejected
> because 2.15.29 in the project-id line, which doesn't match the file
> name for 188.8.131.52
> My surprise is to see that latest accepted file for plain 2.15.37 has
> the same line "2.15.29". How was it accepted?
> Anyway some a little bit wrong is here if we are generating those
> with .29. What do you think?
As far as I remember, I always modify fr.po's version manually. The FTP
in fact does not extract anything else than the pot file (just have a
look at the Dutch file, PO-Revision-Date: 2007-05-29 on the FTP but
PO-Revision-Date: 2011-03-01 in the sources tarball).
I think it comes from the mess I did when submitting a first version of
2.15.37 (on April 21st): the robot was somewhat unable to retrieve the
tarball from download.free.fr and then picked it from audiolinux, so to
say the release built by Graham before I ran po-replace.
That's the reason why I "introduced" a fake version 184.108.40.206 since you
cannot upload a version more than one time.
Hopefully Graham has not released 2.16 in the meantime (consider you
have sent the info to FTP on the 22nd, so to say one week ago!)
This is the very reason why I consider it mandatory to update the
template (lilypond.pot) within every release process, and just
acknowledge the FTP when significant change (running po-replace let' you
know the changes in po files).
What I've done so far is adding a rule to
- only adds a new lilypond.pot to the release-tarball
- use sed to update its headers (first 3 lines, version number,
Report-Msgid-Bugs-To and Content-Type).
But I see many CHARSET errors I would like to understand and try to get
rid of, before proposing anything to devel- (and avoid any shooting from
On another hand, I would like to have "programming-interface.itely" be
part of 2.16.