Re: Looking for a Turkish user of tin

On Mon, Mar 07, 2005 at 12:50:25AM +0200, Deniz Akkus wrote:
> > complete? a saw 72 translated messages (out of ~1200).
> The rest is marked as "fuzzy" as it needs someone to check it over,
> but they are there.

on the first check I saw a lot of untranslated messages marked as
fuzzy (e.g.):

| #: src/attrib.c:587
| #, fuzzy
| msgid "#  quote_chars=STRING (%%s, %%S for initials)\n"
| msgstr "#  quote_chars=STRING (%%s, %%S for initials)\n"

but on the second look I noticed that there are at least as much
fuzzy translations as just copied text.

> > and somewhat outdated as the translation is based on the strings
> > found in tin-1.6.2 but it's  unlikely that there will be another
> > tin-1.6.x release and a couple of strings have changed in the 1.7.xer
> > branch. (The unstable branch (1.7.x) is not 'managed' by the TP, but I
> > could ask them to do that.)
> Well, speaking as a translator and a translation team leader, it is rather 
> frustrating to work on a translation only to find out that it is outdated / 
> no longer to be used etc. 
> 
> TP shows 1.6.2 as the latest/greatest version of tin. If this is not the case, 
> I would really appreciate you updating it. 

tin-1.6.x is the current stable release, tin-1.7.x is not ment to be
used by 'normal' users, it is the unstable development branch (which
currently is very stable, but that might change whenever a new
feature is introduced or some older code is rewritten). it's a moving
traget and thus I thought it might be not the best idea to bother
translators with it as the strings will change more frequently as in
the stable branch.

so if anyone is going to work on the (initial) turkish translation it would
be more wise to use the 1.7.xer tree as it's base instead of starting
with the 1.6.xer tree as it's unlikely that there will be another
1.6.x release. but of course we can take over the po-file to the
1.7.xer tree and update it accordingly (that's what I allready did,
converted it into iso8859-9 and fixed a few minor 'bugs' - but it
might not be worth keeping it).

> > besides from that the tr.po file is in utf-8 which IMHO is not the best
> > choice as it requires a utf-8 able environment to edit it (I usualy
> > don't use a utf-8 environment) - IMHO iso8859-9 would be a better
> > choice (for the moment and if it suffices).
> Gettext in its later versions handles output string conversion so a utf8 
> encoded translation is not a problem for a user in an iso-8859-9 environment.

tin comes with it's 'own' version of gettext (for those systems which
do not have it) - unfortunately it uses a somewhat outdated version
(0.10.40). dunno if that version can handle utf-8 stuff or argument
reordering like in:

| #: src/lang.c:112
| #, c-format
| msgid "%s %d %s in %lu seconds\n"
| msgstr "%4$lu saniyede %1$s %2$d %3$s\n"

but of course users can always install a newer version on thier system and
link tin against it (and whenever someone is willing to update the
included gettext version in tin then we no longer have to worry about
such problems).

> For editing the file, the Turkish language community uses UTF8 rather more 
> than an iso-8859-9 environment these days. 
> 
> If this is a problem for you as a maintainer, we can definitely encode the 
> file in iso-8859-9, but it is not a problem for the translators. 

it's not really a problem, but usualy I try to fix small
bugs/outdates in the translation files - as I don't use a utf-8
environment I can not do that if the file is in utf-8; this would
lead to some more work for the translators (no problem for me ,-).

urs

p.s.: tin-users is usualy not ment for technical discussions (I don't
      care, but other subscribers might do), so we should take (at
      lest the technical side) over to tin-dev (or if you don't like
      to subscribe another (low traffic) mailing list) or private
      mail.