Very important notice to all developers landing code for the 1.1 release: We need to reduce the string changes to *zero* by June 30. That's 3 days from now.
* If you own the bugs in the searches below, please try to land ASAP. * If you think there's *any* chance that your bugs will have string changes, flag them immediately. * If there's any possible way you can *not* make any planned string changes, please do everything you can to make that so. This release is shipping more locales that 1.0.1, and has a large number of string changes and additions. We are still landing too many changes, and continuing to do so puts our release at risk. ----- Original Message ----- From: "Axel Hecht" <[email protected]> To: "Dietrich Ayala" <[email protected]> Cc: "Joe Cheng" <[email protected]>, "John O'Duinn" <[email protected]>, "Wilfred Mathanaraj" <[email protected]>, [email protected], "Kevin Hu" <[email protected]>, "l10n-drivers" <[email protected]> Sent: Thursday, June 27, 2013 7:59:58 PM Subject: Re: L10n status for 1.1 Hi, quick recap on the context: The keyword late-l10n is really best used to monitor the backlog of work that not only needs to be done, but also needs to be localized. That is, it needs to be done on a tighter timeframe. After mega-triage on the 19th, that were this 7 bugs . We had a l10n status meeting that day, too, and Alex and folks expected those to be done by the end of this month. We're down by 3 on that list. The query I ran back then shows 6 bugs right now. Which means that we got the bug count down by 1 bug in a week. For us to ship a localized product, we really need everyone to help to get that buglist out of the way. Doesn't matter much if they're fixed, wontfixed, fixed by a branch patch that doesn't do the last bit of eyecandy and thus doesn't change strings. And obviously, stop putting new bugs on that list. Anything that you can do to get engineering focus on those bugs, and the ones blocking them, would help a lot. Also, educate the triage folks. The cost of a late-l10n bug is much higher than of another bug. Explain that adding the late-l10n keyword doesn't make a new blocker OK, all it can do is give visibility to how we're doing in the schedule. PS, the MMI strings I just uplifted added 75 strings. According to ferjm, not all of them might be final. Axel On 6/27/13 11:52 AM, Dietrich Ayala wrote: That's great, thanks! How can we help identifying why strings are coming in late? Are people not adding late-l10n to their bugs? If you have recommendations on how we can help, we'll gladly try to improve the situation. On Thu, Jun 27, 2013 at 2:46 AM, Axel Hecht < [email protected] > wrote: <blockquote> Hi Dietrich, the good news is, we pass that metric now, for all 1.1 locales but hu, ro, sr. We have 2500 strings in total inside gaia, which means that anything missing less than 250 strings would be OK by that metric. Note, the percentages you see on the dashboards are different, as there's some level of strings in each language that are literally the same as in English, but perfectly fine. That limit seems to max around 90%, and depends on language. pt-BR has 245 missing strings, so it's passing that metric now. Given that we're still expecting a bunch of string to come in to en-US, it could drop beneath by July 8. Looking that the MMI strings will show up later today, maybe today. Outside of that metric are Hungarian, Romanian, Serbian. So, we're passing right now, and we're going to get better. I would like to add that we're continuing to add string for bugs that are not late-l10n, that is, weren't in scope after mega-triage. The MMI strings are also just coming in, I'll add them to the dashboard later today. We're mentioning earlier dates at the same time. I'll focus on July 15. I am pestering people to land early, but I'm not successful. I'm getting replies, which is good, but not as many landings as I'm hoping for. Axel On 6/27/13 11:12 AM, Dietrich Ayala wrote: <blockquote> Hey Axel! I have a question for you: To enter IOT testing, it is best if En, Es & Pt-Br are ~90% by July 8. Is that a possibility? It's difficult to tell with status of good below. Thanks, Dietrich <blockquote> Looping in Kevin Hu, too. No responses on this so far, Kevin, so you didn't miss anything but the initial mail. Axel On 6/25/13 9:11 PM, Axel Hecht wrote: <blockquote> Hi, here's the current status of localizations for 1.1: Locale Status Build cs Czech good dev de German good dev el Greek started last week, needs loads a testing, mostly translated dev es Spanish good dev, 1.0.1 hr Croatian on track, needs more landings dev hu Hungarian on track, needs more landings dev nl Dutch good dev pl Polish good dev, 1.0.1 pt-BR Portuguese BR on track, needs work, target July 15 dev, 1.0.1 ro Romanian needs feedback, landing problems dev ru Russian good dev sk Slovak needs work dev sr-Cyrl Serbian needs work, unlanded work exists dev tr Turkish good dev bg Bulgarian not started, no repos. Rebuilding community with interest ca Catalan on track, no device testing ja Japanese on track, no device testing. Fonts? Keyboard code exists mk Macedonian not started, no community. Desktop is hardly alive ms Malay not started th Thai on track, no device testing. Fonts? Keyboard? Sources: https://github.com/mozilla-b2g/gaia/blob/v1-train/locales/languages_dev.json is what's on the nightly builds. https://l10n.mozilla.org/shipping/dashboard?tree=gaia is the status we're tracking, https://l10n.mozilla.org/shipping/dashboard?tree=gaia-community are other locales. https://intranet.mozilla.org/Program_Management/Firefox_OS/Localization for open items, recent edits haven't been synchronized. Which we'll do now. https://l10n.etherpad.mozilla.org/fxos-new-locale-checklist has a start of a list of things we need to do for new locales. I didn't dive into India, too many languages to be useful right now. Pakistan is hard, Urdu would be a RTL locale, not sure if we're even close. Let's discuss Axel _______________________________________________ L10n-drivers mailing list [email protected] https://mail.mozilla.org/listinfo/l10n-drivers </blockquote> </blockquote> _______________________________________________ L10n-drivers mailing list [email protected] https://mail.mozilla.org/listinfo/l10n-drivers </blockquote> </blockquote> _______________________________________________ dev-b2g mailing list [email protected] https://lists.mozilla.org/listinfo/dev-b2g
