Axel Hecht Mozilla in Your Language

I’ve spent the last few days catching up on years of archive of translate-i18n. It’s been an interesting read so far, the challenges we face are the same in other places, and are pretty independent of the used technology. Given the recent 50-2-100 discussion in .l10n and .planning, I found two threads from April this year particularily interesting. Morale and Draft of awareness-raising email to maintainers. The first is somewhat of a corresponding thread to 50-2-100 on their side, and the latter is a resulting inquiry for project maintainers to see if they do the right thing for localizers. As the survey is kind-of fun and adds some non-Mozilla context, I took the freedom to fill it out for you.

1. Do you freeze your strings at least two weeks before release?
   (a) Yes
   (b) No
   (c) What strings?
2. Do you send in a pre-release POT file, at least two weeks before
   release?
   (a) Yes
   (b) No
   (c) POT? Not. What rot!
   … as tempting as (c) is :-)
3. Do you fill in the Report-Msgid-Bugs-To address in the PO file
   header?
   (a) Yes
   (b) No
   (c) I’m in hiding.
4. Do you supply contextual comments for complex strings?
   (a) Yes
   (b) No
   (c) I don’t understand them either.

The message? No, we’re not done, by far. But we’re addressing the right targets, and we’re improving in key areas that are of concern to localizers, inside and outside of the Mozilla project.

During the l10n session today we came up with an request from dynamis for him to be able to not have a string argument show up in a formatted l10n string.

We came up with the idea to specify zero width in the format, but it’s actually precision 0.

foo=does "%1$.0S" work?

as foo.properties does actually yield to

js> sb.formatStringFromName('foo',['Firefox'],1)

does "" work?

So here goes your nifty l10n tip of the day.

Not speaking Punjabi actually made me speak Chinese. Well, not really.

The lesson to take away is, you may not be able to speak a language, but looking at a patch on a mac with Chinese fonts sure does catch some bugs. Looking at the bugzilla diff pages in UTF-8 encoding is a good exercise when approving patches to unkown languages.