[Localization] codec can't decode byte (was: Localization efforts needed)

[F Wolff]

Op Ma, 2009-09-21 om 12:24 -0400 skryf Alexander Dupuy:
> Chris writes:
> > Please log onto the Pootle server and reviewing your favorite language(s).
> >   
> 
> I was attempting to update my Projects to include the 0.84 release 
> project (although I'm not 100% certain that these are the ones you are 
> talking about, as opposed to the the unversioned Sucrose, Fructose etc.) 
> and I got the following unfortunately familiar error:
> 
> Error 'ascii' codec can't decode byte 0xc3 in position 4: ordinal not in 
> range(128)
> 
> Sayamindu wrote (in July 2008):
> > Apologies for the delay in fixing this - apparently a user had
> > registered with a non ASCII username which was choking Pootle. Had to
> > find it out and disable it. It should work properly now.
> 
> There is a Pootle bug about this 
> (http://bugs.locamotion.org/show_bug.cgi?id=954) but it doesn't have 
> much in the way of details or anything...  I'm cc'ing Friedel Wolff, who 
> was assigned that ticket, in case he has any suggestions or comments (or 
> in case you can provide him with more details on this problem).
> 
> 
> Sayamindu later (in August 2008) wrote:
> > For cases like yours, (and Bastien's), please mail
> > translate at rt.laptop.org to ensure that this does not get lost in my
> > mailbox. :-)
> 
> I am cc'ing that address as well, although I'm not sure that it is 
> relevant anymore or even used now that the OLPC/Sugar Labs Pootle 
> instance has moved over to Sugar Labs.
> 
> @alex


Hallo everybody, and sorry for my slow reply after some time offline.

I have also seen this issue, although it is not quite as straight
forward as a non-ASCII username. I have definitely at times tested
non-ASCII usernames and it works, but it seems that Alex might have
found some more details which he added to this bug in our (currently
offline) bugtracker:
http://bugs.locamotion.org/show_bug.cgi?id=954

I am quite sure that the issue is with jToolkit that is handling the
user accounts, and this has all been replaced entirely with Django for
the upcoming version. As soon as our testing server is back online, we
can quickly test to ensure that it is working well, and hopefully the
issue will simply disappear with the next version.

I have at times found that a restart _seemed_ to fix the issue with the
non-ASCII usernames continuing to work afterwards. I realise that this
is not quite satisfactory, so hopefully we can eliminate this constraint
soon.

Keep well
Friedel


--
Recently on my blog:
http://translate.org.za/blogs/friedel/en/content/support-afrikaans-google-translate