110 replies to this topic

[ranvis]

mozjonathan,
I'm wondering if the string indexing.p5 in pages.dtd;
"You're in one of the two situations below." should be "You're in one of the two situations above." or am I missing something?

[Sonickydon]

At first, I'd like to apologize for taking me so long to complete the localization.
With all the Firefox 4 -release-frenzy I had to give priority to updates of already public stuff and let not-yet-released extensions behind.
Of course this will not happen in the future (once the locale gets public).

To more serious stuff now:
It might just be the beer talking right now but it seems i can't test the localization.
I've added the locale to install rdf, chrome manifest, added the files in the jar -in the locale folder under the locale code named folder, tried with the version of the xpi from babelzilla and the one from amo but it doesn't seem to work (?)
I get a parsing error when i try to open the properties dialog, and the error console reports one undefined entity (no missing strings in the files though) and a message saying that it could not read the chrome.manifest file.
Now, the extension id that the error console reports is {972ce4c6-7e08-4474-a285-3208198ce6fd} while the one in the install.rdf is {3550f703-e582-4d05-9a08-453d09bdfdc6}, could that be the problem?

[markh]

Now, the extension id that the error console reports is {972ce4c6-7e08-4474-a285-3208198ce6fd} while the one in the install.rdf is {3550f703-e582-4d05-9a08-453d09bdfdc6}, could that be the problem?

The latter is the id of Thunderbird itself. {972ce4c6-7e08-4474-a285-3208198ce6fd} appears to be some sort of default extension or theme.

[Sonickydon]

I can't see any missing strings in my files, the only thing I can assume is that something goes on with the not-yet updated localization of Miramar itself for my locale since it works for everyone else here.

[mozjonathan]

mozjonathan,
I'm wondering if the string indexing.p5 in pages.dtd;
"You're in one of the two situations below." should be "You're in one of the two situations above." or am I missing something?

That's right. This will be fixed before the release.

[mozjonathan]

Hi all,

The following descriptions for install.rdf are missing:
- el
- hu-HU
- pt-BR
- sv-SE

Thanks!

jonathan

[mozjonathan]

I've changed some strings, so most translations will need a quick update. Here's a build that you can use to try out your locales:

http://jonathan.xulforum.org/files/gcv-nig...2214-master.xpi

Cheers,

jonathan

[Sonickydon]

el released (finally!), a big thank you to Goofy and mozjonathan for their help in resolving the testing issues.

Localized description for the instal.rdf:

<em:localized>
<Description>
<em:locale>el</em:locale>
<em:name>Thunderbird Conversations</em:name>
<em:description>Προσφέρει εμφάνιση μηνυμάτων σε μορφή συνομιλίας στον Thunderbird!.</em:description>
<em:translator>Γιώργος Φιωτάκης</em:translator>
<em:creator>Jonathan Protzenko</em:creator>
<em:homepageURL>http://github.com/protz/GMail-Conversation-View</em:homepageURL>
</Description>
</em:localized>

@ Translators and mozjonathan:
While checking the install.rdf file I noticed that almost all (except he-IL) of the localized descriptions still mention Gmail while that string has changed.

[ranvis]

Here's ja locale description:
<em:description>返信メールが一目でわかるスレッドビューをあなたのThund
erbirdに!</em:description>

BTW, sepComma in message.properties requires explicit space at the end, whereas Babelzilla automatically trims it.
I tried using U+2004 THREE-PER-EM-SPACE(" ") as an alternative. Any better ideas?

[Goofy]

Hello mozjonathan. I suggest to push the necessary space separator into the code part, to me it can be fixed without translators trying to insert space. Here is a tentative patch (probably not sufficient/valid but this is to convey the suggestion)

http://dl.dropbox.com/u/20866027/spaceSepa...on04162011.diff

[ranvis]

OK, I read en-US/message.properties and learned that "\u0020" solve this trailing space issue. I was even trying to file a feature request...
mozjonathan, please make it written down in translation comment

[mozjonathan]

I've integrated the greek description, thanks!

Concerning the space issue, I don't want to hardcode the space into the code, because I think it is up to the translator to decide whether a space is needed after and/or before the separator. In Chinese, for instance, the half-comma requires no space after, so always adding a space after &sepComma would be wrong. Here are the various issues:
- the Gecko parser for properties files strips any space at the beginning and at the end of the string... using \u0020 at the end of string solves this issue however,
- Babezilla thinks it's smart to replace \u0020 by the corresponding space character, and this makes the whole thing fail. At least when I upload a file that contains \u0020, it is replaced by a regular space...

Goofy, do you think you could get someone to look into point #2? It's breaking the whole think, and I don't want to have the translators use some weird space such as three-per-em space like ranvis cleverly suggested (other spaces for you to pick: http://www.fileformat.info/info/unicode/ca...ry/Zs/list.htm). Ranvis, is the \u0020 preserved after you download the localized file?

[Goofy]

[snip]
- Babezilla thinks it's smart to replace \u0020 by the corresponding space character, and this makes the whole thing fail. At least when I upload a file that contains \u0020, it is replaced by a regular space...

Goofy, do you think you could get someone to look into point #2? It's breaking the whole think, and I don't want to have the translators use some weird space such as three-per-em space like ranvis cleverly suggested (other spaces for you to pick: http://www.fileformat.info/info/unicode/ca...ry/Zs/list.htm). Ranvis, is the \u0020 preserved after you download the localized file?

Sure we can look into this, but i am afraid this conversion was necessary/requested for other cases, so it is possible if we fix this as per your request that there will be other files broken with preservation of escaped unicode space char I will let you know.

[ranvis]

Ranvis, is the \u0020 preserved after you download the localized file?

"Download" > "All locales as .tar" > "strings replaced/blank/skipped" works.
Don't know about new downloading feature for extension developers.

[mozjonathan]

Hi everyone,

I've uploaded a new version which is what will become Thunderbird Conversations 2.0beta1; I expect to release it at the beginning of next week. I'll release another beta with all locales activated if some people are not ready by then.

Thanks!

jonathan

[Lakrits]

In template.properties we're supposed to translate a single preposition, "at". That preposition cannot readily be translated into Swedish, since it's replaced by several other Swedish prepositions depending on the context. It can be compared to the handling of verbs in different languages. In English, French and Swedish, the verb "to be" is conjugated so differently that single translations of it will result in incompatible grammatical errors. For example, in present tense:

CODE

English (to be):

I     | am
------|--------
You   | are
------|--------
He    |
She   | is
It    |
------|--------
We    |
You   | are
They  |
---------------

French (d'être):

Je    | suis
------|--------
Tu    | es
------|--------
Il    |
Elle  | est
On    |
------|--------
Nous  | sommes
------|--------
Vous  | êtes
------|--------
Ils   | sont
Elles |
---------------

Swedish (att vara):

Jag   |
Du    |
Han   |
Hon   |
Den   | är
Det   |
Vi    |
Ni    |
De    |
---------------

For the preposition "at", here are some examples:

CODE

English         Swedish

at once         genast

at hand         till hands/till handa

at click        vid klick(ning)

at Sunday       pÃ¥ söndag

at 4 p.m.       klockan 16:00

For this, and other, translations of the same kind, I will need commentary lines with examples showing me what you are looking for. If this preposition is used in more places than one, a single translation might not be enough.

/Lakrits

[mozjonathan]

Hi Lakrits,

Sorry, I should've thought earlier that this warranted some explanations. In some cases, we will detect what I call a "Bugzilla" thread, which is basically a thread of emails generated by the bugzilla system (see http://bugzilla.mozilla.org for an example). In that case, instead of showing, in the top of the message box, "person A to person B", I will show "person who commented on the bug AT http://bugzilla.whatever.com".

Does that shed some light upon the matter?

Thanks

jonathan

[Lakrits]

Yes, it does.

I guessed that you meant something like that and wrote the appropriate translation in the file, but you never know…

/Mikael

[mozjonathan]

Great. I've added a comment near the line, I'll try to reupload a new version shortly so that other localizers can see the comment. If you spot any other ambiguous strings, please do let me know!

Cheers,

jonathan

[mozjonathan]

Hi everyone,

I've created a dev build with all locales that are at 100% enabled. http://jonathan.xulforum.org/files/gcv-nig...2333-master.xpi

Feel free to try it out. There's still the issue of sepAnd and sepComma for which Babelzilla (stupidly) converts the \u0020 into a regular space, but I'll personally verify that no final version is shipped with the issue, so don't worry about that (I'm trying to bribe the Babelzilla devs into fixing it) :-)

jonahtan