discussing in detail my position and my problems with wordlist.po

[Arafat Medini <lumina at silverpen dot de>]

As I've done some work today on gconf.po I want to show some of the
examples why I think the wordlist needs far further improvement:

1)the string in gconf says:
"Recieved·signal·%d,·dumping·core.·Please·report·a·GConf·bug."

And this is what the wordlist gives me for dump/dumping:
Dumping الإغراق
Dump النفاية

are these words really useful for this string?!
But are these translations false? also no!

2)randomly choosen words:

Quit متحرر

Close الإنتهاء

Shut أغلق

Sync المصاحبة


3)a very hard word "flush" cz of many double meanings:
the string says"Failed·to·flush·client·add·to·saved·state·file:·%s"

the wordlist gives me:
Flush الإحمرار

which doesn't have ANY relation to the string needless to say that even
the standard translation of flush (get rid) also doesn't work here
because of the grammatic structure.
Flush here means to spread over.

some words where we need concensus:

Subdirectories أدلّة ثانوية
 what about this:
دلائل
محوية

I didn't write this but I found it even better then the one in the
wordlist it feels more accurate, righter in the gconf file.
(those are only some examples I had the problem with maybe every third
word or more...)

So as you can see we have so many problems with the wordlist  with
RESPECT to it's achievement, that it's far away of making it a standard
for DE translation. simply putting thousands of words in a file and then
saying this is the standard is not the right path.
We need a complete overwhaul of how the wordlist responds to needs like
DE translation projects have.
some ideas:
- multiple choice
- more accurate words
- maybe branch wordlist po in multiple wordlists ??
- meetings btw us and the wordlist ppl are also welcomed.

I can't accept making translation for a DE which will be used by Novell,
SuSE, IBM and Redhat with sticking to something I am not convinced of.

Though I want to repeat something: the wordlist po is a great work done
by great ppl, I respect the project and if it fulfills my needs I'll be
the first one to use it!

But for now it doesn't fulfill my needs at all. From the words I
searched for, only general terms were accurate, if the file fails to
give me a good translation even for "quit" so what about really complex
debug strings? Or system sensitive strings like changing keys or very
complex grammar? which needs completely different words then a one to
one translation?

So that's why I am not in favour of using the wordlist as a standard.

Standard Yes, but standard at any price? no.

It would really help to make the wordlist better but I don't have time.
I mean gnome is so big, if I really had time I would discuss with the
wordlist ppl all the points I don't like to make them work on it, but
that's not possible for me.
And on the other hand time precess I want to get gnome at least clean
and well integrated in itself before the big companies roll out their
distros. It is a shame that we don't have mature good Arabic support for
Linux which will be the next big thing.

So my stance is as said, am I in favour of a standard? hell yes!
Does wordlist fulfill my needs ? Now NO.
 What do we need?
Discussion!
am I against discussion? hell no!
Al-rashedeen I love you! (I am not gay,btw...)

that's all what I've got to say!

cheers
Arafat