I don't think we felt blamed (I definitely didn't), and the support for l10n has been great, especially in terms of responses and support :-)

We have pain points on both sides (e.g. you have nothing to do with the Pootle issue) and we're fixing holes in the road as we go. From my point of view it's inevitable for a v1, and so far it's going even better than I expected.

The was merely an explanation for the scarcity of l10n bugs so far, things will change from next week.

Francesco

To be clear, I did not intend to blame the L10N community in any kind of way. We are extremely happy and impressed with the L10N help for this project and everyone has been doing an absolute great job dealing with this project. You all rock. I am not joking.

If there is anyone to blame then it is me messing up those scripts and not including all the strings in the builds.

I think one of the problems of our current Test Flight program is that our test audience is not diverse enough. It would be ideal if we we have 4 or 5 people testing in each supported locale. But we don't. I think we currently have a small subset of the L10N people and then next to that mostly english and german testers. So that is not optimal. For the (internal) Aurora builds this is even worse I think.

I'm sure we can find more testers but I don't know how to fix this in the short term.

 S.


On Thu, Jul 2, 2015 at 4:05 PM, Francesco Lodolo [:flod] <[email protected] <mailto:[email protected]>> wrote:

    Hi Stefan,
    to answer your question "I'm surprised nobody really noticed.", it
    goes back to the discussion we had in Whistler: so far l10n builds
    have been useful up to a certain point.

    I have extra strings localized in Italian, but they're ignored
    because the English reference file doesn't have them. Therefore I
    can't safely report bugs for missing translations.

    Locales working on Pootle are not updated automatically (yet).
    Again, a missing string in the device might be due to Pootle being
    behind with the export.

    Luckily Test Flight builds will get a lot more useful starting
    from this Friday, with strings stabilizing and a clearer process
    for string exports and SVN updates.

    Francesco


    2015-07-02 21:52 GMT+02:00 Stefan Arentz <[email protected]
    <mailto:[email protected]>>:

        I think we are pretty close. I don't think we have any
        outstanding strings at this point.

        I have looked at many l10n bugs and resolved almost all of them.

        L10N related commits/changes:

        Fixes 1163117 - Copy for on-boarding
        
<https://github.com/mozilla/firefox-ios/commit/cef4c9d6425d5371092fbeb95156c831009a7d72>


        Fixes 1172111 - Copy for remote tabs panel
        
<https://github.com/mozilla/firefox-ios/commit/f428f44068300c3548195c9ae63432fe3c0505fd>


        Fixes 1168402 - Copy for Reading List empty-state
        
<https://github.com/mozilla/firefox-ios/commit/70761d5711e43507dd787df8d5a6a659a5a0d55f>


        Fixes 1168400 - Copy for Send Tab extension empty state
        
<https://github.com/mozilla/firefox-ios/commit/77bb893481ff28b8aeae2e3d946c11336aede32f>

        Fixes 1151995 - [l10n] Fix string consistency: "Tap to reload
        page"
        
<https://github.com/mozilla/firefox-ios/commit/0144f25f83061bdb8ce417c873fb9ee88582224b>

        Fixes 1153333 - Missing notes for NSLocalizedString
        
<https://github.com/mozilla/firefox-ios/commit/ad8108e68f367adbfbf911278c624c3dc90c276b>

        Fixes 1163151 - Firefox for iOS: verify usage of newslines in
        strings
        
<https://github.com/mozilla/firefox-ios/commit/0f491a644b4b32a52ea170d1137b93caefe312dc>


        Fixes 1179863 - Copy for Send Tab not logged in state
        
<https://github.com/mozilla/firefox-ios/commit/503894866f877d0155034797ee56cf20bf933ef2>


        Fixes 1175887 - Not all string tables are properly converted
        to .stri…
        
<https://github.com/mozilla/firefox-ios/commit/6e7d3ee3567ad3b3fbcf29e07c267ec48ef54731>


        The last one is worth mentioning: a number of strings were not
        actually included in the build. I'm surprised nobody really
        noticed. I think that shows that we need to figure out a way
        to do more automated testing. For example with screenshots
        taken of every possible piece of UI that we have.

         S.


        On Tue, Jun 30, 2015 at 10:53 PM, Stefan Arentz
        <[email protected] <mailto:[email protected]>> wrote:

            As discussed last week in Whistler, we are aiming for L10N
            string freeze by the end of this week.

            I have browsed through the app sources and I could not
            find any static (non-NSLocalizedString) strings. Please
            prove me wrong.

            We do have a bunch of strings that need a quick review
            before they can be localized in their final form:

            
https://bugzilla.mozilla.org/showdependencytree.cgi?id=1168395&hide_resolved=1

            I'll hope to get those finalized on Thursday.

            My questions to you are:

            * If you know we are going to add new functionality with
            new strings in the coming days/week, please add them as
            soon as possible. Or discuss so that we can find a way to
            get them translated before the code lands.

            * If you know of more strings in the app that need review,
            please file a bug and make it block 1168395.

            What did I miss?

             S.



        _______________________________________________
        mobile-firefox-dev mailing list
        [email protected]
        <mailto:[email protected]>
        https://mail.mozilla.org/listinfo/mobile-firefox-dev




_______________________________________________
mobile-firefox-dev mailing list
[email protected]
https://mail.mozilla.org/listinfo/mobile-firefox-dev

Reply via email to