PO and POT files for Sympa project

Tue May 10 10:01:56 CDT 2005

On Tue, 2005-05-10 at 11:52 +0200, Olivier Salaün - CRU wrote:
> 
> > Well, technically, *now* are editable. I found a problem yesterday that
> > prevent me to do the update and It was fixed a couple of hours ago.
> > 
> >   
> > > There's one thing that annoy me though : The English, French, Italian and Hungarian translations are almost 100% completed but they don't appear on the Rosetta web interface. Do you have a way to take existing PO files into account ? Here they are : 
> > > http://sourcesup.cru.fr/cgi/viewcvs.cgi/sympa/po/?cvsroot=sympa
> > >     
> > As I said, I was not able to do the full import yesterday but it should be done now.
> >   
> Everything is in place now, thank you.
> I'm just wondering about how "fuzzy" entries appear on the Rosetta web
> interface. 
> For most of our PO files, fuzzy entries should not appear at all.

The .po files you pointed me at have lots of fuzzy flags there, Rosetta
just shows you what it got.

> 
> Another question is : how do you cope with different characters
> encoding ? Most of our PO files are not UTF-8, mainly because UTF-8 is
> not yet well supported by mail clients. Do you consider everything
> UTF-8 ? Do you recode characters to have an UTF-8 web form ?

Rosetta works internally with UTF-8 strings, but accepts other inputs
from the .po imports and the export is done with the encoding specified
in the .po header. If it does not work that way, it's a bug.

For new .po files you will need to upload a .po file into Rosetta with
the wanted encoding because we don't have a way to select it atm (could
you add a bug report about it or add an entry at
http://www.ubuntu.com/wiki/RosettaWishList ? ). By default, Rosetta
exports new .po files as UTF-8.

> > Btw, many of the .po files does not have an owner so Rosetta, by default, put the importer (me) as the owner of those translations,
> > please, open bug reports with the name and email address of the right owner for those translations and I will change it.
> >   
> Strange : you should have authors name, at least in fr.po, hu.po,
> it.po, en_US.po.
> They appear as follows in the PO file :
>         msgid ""
>         msgstr ""
>         "Project-Id-Version: fr\n"
>         "POT-Creation-Date: 2002-07-16 17:27+0800\n"
>         "PO-Revision-Date: 2005-05-03 17:50+0200\n"
>         "Last-Translator: Olivier Salaun <olivier.salaun at cru.fr>\n"
>         "Language-Team: Français\n"
>         "MIME-Version: 1.0\n"
>         "Content-Type: text/plain; charset=ISO-8859-1\n"
>         "Content-Transfer-Encoding: 8bit\n"
> Where do your scripts expect the author to be defined ?

We parse the Last-Translator field to get that information

Right, fr.po, hu.po and it.po have the right value, en_US.po has the
default value, so Rosetta used the importer info (me).

> > > Other minor concerns : how often should we submit updates of our POT file ? Is the integration in Rosetta an automatic process ?
> > >     
> > As often as you want your translations get the updates.
> > 
> > If you are not adding new strings or changing existing ones, you don't need to update the .pot file.
> > 
> > About the automatic process, no, we don't have yet defined a way to automatically fetch .pot and .po files from upstream projects like yours. We only do it with .po and .pot files that come from Ubuntu's archive because it's the same process for all .deb packages. We are working on a process that will let us do a kind of automatic import for upstream but we cannot give you a date for it, sorry.
> >   
> Ok.
> Another question that came to my mind : when we trigger an update of
> POT and PO files on Rosetta, do you replace the existing ones with new
> ones or do you perform a merge. In other words, do you expect us to
> integrate the updates from the Rosetta users before we submit new
> files ? 

The new .pot file overwrites the old one in Rosetta.

The new .po files are merged. By default, changes from Rosetta have
preference over the old translations (so we don't lose the changes done
from Rosetta) but if we detect that a translation was done in the .po
file and in Rosetta, usually, the .po one has preference.

If you plan to use Rosetta as your main way to get translations, I think
the best workflow could be that you upload from time to time an
updated .pot file and translators translate from the website or get
the .po files from Rosetta, translate offline and upload them into
Rosetta later and when you are going to do a release, you can get
all .po files from Rosetta and add them into you CVS (or also from time
to time so cvs snapshots get translation updates, as you prefer).

At this moment you need to get every po file one by one, but as part of
our work in next two months we will provide a way to get a tarball with
all .po files inside.

> 
> I think it's time for us to do some advertisement for the Rosetta
> translation interface in our sympa-translation mailing list.
> 
> Thank you Carlos.
> 

No, thank you for use Rosetta!

Cheers.

> 
-- 
Carlos Perelló Marín
Ubuntu Hoary (PowerPC)  => http://www.ubuntulinux.org
mailto:carlos.perello at canonical.com
http://carlos.pemas.net
Valencia - Spain
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: This is a digitally signed message part
Url : http://lists.ubuntu.com/archives/rosetta-users/attachments/20050510/b28ab89a/attachment.pgp