On Mon, 2007-01-01 at 20:34 +0200, F Wolff wrote:
> Thank you, Richard.
> Just a few questions:
> * Are fuzzy messages currently used? I saw a strange link on one page
> that seems to be because of a fuzzy message being used. Mostly not a
> big issue as I'm currently updating the translation anyway.
We use a custom message merge script that Dominic wrote, which uses some
fuzzy translations but has a stricter definition of fuzzy than the
normal msgmerge (to try and avoid spurious matches).
> * Several translatable sentences are constructed without the full stop
> and then the full stop is added in the code. Is this safe to do? Won't
> some languages require control over their ending punctuation? I _think=
> Khmer for example does not use a full stop at the end of sentences. Thi=
> even makes it a bit difficult for somebody who would need the full stop=
> since we might add it under the assumption that the English is wrong.
> Then we will end with two full stops.
Seems fair to me. I can't see any reason why we shouldn't put the full
stop inside the translated string.
> * The message currently at download/plugins.php:20 ("Audacity has
> built-in support...") does not have a full stop, and, in contrast with
> the ones mentioned before, actually seems to need it.
I agree, that needs a full stop, so I've added it (inside the translated
string) to the site. I went through and checked the fuzzy translations
that produced - it turns out a majority already had a full stops! I've
fixed all the fuzzys except the ones in non-eurpoean scripts where I
didn't know what to.
> * Not everything on community/developers is internationalised. See the
> section on CVS for example.
I can't see anything except the actual CVS checkout URLs which isn't
translated - am I missing something?
> * The news entries (notably the current front page news, "Audacity
> 1.3.2 and 1.2.5 Released") are extremely long. Isn't it possible to cho=
> it up into a few smaller chunks? Currently it is difficult or
> impossible to use translation memories, or to reuse this translation
> elsewhere. It is a bit frustrating if I can't even complete the
> translation of _one_ message in a five minute break :-)
Not being a translator, I've never been sure which way to go with things
like this. Is breaking down into short chunks (a sentance? a paragraph?)
more helpful than the context provided by keeping stuff together? I take
it from what you say that the former is better?
> * Ditto for some of the release notes.
> * Talking about release notes, perhaps we can include the fact that th=
> Afrikaans translation made it into 1.2.5?
Oops. Release notes are a bit of a faff because they are usually written
up as a text file in the source code CVS, and then copy-and-pasted to
the website. As a result I'm unwilling to mess with the 1.2.5 notes
(because they are in all the tarballs etc). We can try and do better
with the next round, at least in terms of getting them translatable (I
hope we get it right, as well ...).=20
> * I'm just wondering about the use of smart quotes (=E2=80=9CBla=E2=80=
=9D instead of
> "Bla"): In cases where people are using some non-Latin font and a
> message is untranslated, don't we stand the danger of the quotes not
> being displayed properly in the English fallback? Furthermore I find it
> a bit cumbersome to type, but I guess I'm just lazy ;-)
Do we have any smart quotes on the site? I can't think how I'd put one
into a page, but it's quite likely there are some I just haven't seen.
Can you point me at an example?
> * Is it possible to localise the Google form? (include/header.inc.php)
> It would even be nice if we are able give the Google URL to the
> localised version, but I guess this is not so important.
> * Ditto for the alt text for the Store image (line 73, alt=3D"Audacity
I can do that one, it will show up as the string for translation in
include/header.inc.php (because that's where the string actually lives).
It should be in the .po files / pot file on CVS now.