Re: a little start at improving PSPP translatability

[unknown-1]

Internationalization is always fun. :-) and I think this is more
internationalization as translation.

A few thoughts about the remarks:

2009/7/16 John Darrington <address@hidden>:
> Regarding commit "Avoid translating strings that don't need translation",
> I think your assertion that "Strings such as "%d" do not need translation"
> is incorrect.
>
> This is apparantly important for languages such as Arabic.  See
> http://www.gnu.org/software/gettext/manual/gettext.html#index-Arabic-digits-1137

Interesting, never thought about that. But off course %d is not
meaning full, a translator can't do anything with that without further
information or (s)he has to look in the code. So I would suggest to
add some comment for the translator what to do with it.

Another point, in Latin text we write from left to right. In Arabic
from right to left. I have really no idea with happens if you
translate PSPP in Arabic. Anybody on the list who knows this or can
test it?

> Also, I believe that the ampersand character doesn't hold the same
> semantics in every language, so has to be translated too.

I am not sure if I understand, you mean %d could be different in say
Cyrillic? Hmmm. Anybody on the list knows?

> Somewhere else I have  read (I can't remember where) that arabic
> translators also like to introduce special "punctuation" (in the
> form of a long stroke) between certain words, depending on their
> context.  So even strings like "%s %s" need to be translated too.

Here my thoughts are the same as with %d. A remark in the .po file
would be very useful.


Another question is about N. It is a translatable string now. Is it
common in Cyrillic to translate it to H ?

Have fun