Closed Bug 662746 Opened 9 years ago Closed 8 years ago

[my] [Fx] New localization: Burmese

Categories

(Mozilla Localizations :: Registration & Management, defect)

defect
Not set

Tracking

(Not tracked)

RESOLVED FIXED

People

(Reporter: arky, Unassigned)

References

Details

Attachments

(2 files, 2 obsolete files)

User-Agent:       Mozilla/5.0 (Macintosh; Intel Mac OS X 10.5; rv:2.0.1) Gecko/20100101 Firefox/4.0.1
Build Identifier: 


Code: My-MM
Team Page: https://wiki.mozilla.org/L10n:Teams:my
Project Managed Here: https://l10n.mozilla.org/narro/narro_project_list.php?l=my-MM&f=0
Firefox Initial Repo: https://bitbucket.org/arky/my-mm


Reproducible: Always
Duplicate of this bug: 631921
Summary: Registration of Burmese (my) Team → [my] [Fx] New localization: Burmese
OS: Mac OS X → All
Hardware: x86 → All
The Firefox Aurora has reached 80% completion. Request review for creation of offical locale.
Axel Any update on this bug?
We'll need an attachment of the source and a review request on me, that's how I find bugs that actually need a review from me.
Attached file Needs code review (obsolete) —
Needs code review
Attachment #569340 - Flags: review+
Axel how do I tag this bug as needs review from you ?
Set the review flag to "?" and add me as requestee, http://www.bugzilla.org/docs/4.0/en/html/flags.html
Attachment #569340 - Flags: review+ → review?(l10n)
Comment on attachment 569340 [details]
Needs code review

Sorry for the lag, there's been tons of stuff going on. In particular as I'm going to r- this one.

The translation is a good start, but compare-locales finds 28 errors now, and that's a bit much. Should be long for you to fix those, and I promise the next review will be quicker.

Also, could you please switch of the contributor exports in narro? They're violiating our code formatting guidlines, in particular, the license block is wrong.

I'll also go through the full monty of https://wiki.mozilla.org/L10n:Review_notes, so the next review shouldn't have to find new things.

So, the new version of compare-locales find 28 errors, and 9 warnings, some of which are errors, too:

my
  browser
    chrome
      browser
        aboutDialog.dtd
            ERROR: Unparsed content "<!-- LOCALIZATION NOTE (community.mozillaLink): This is a link title that links to http://www.mozilla.org/. -->
<!ENTITY community.mozillaLink      ""မော်ဇီလာကွန်မြူနတီ"">" at 288-459
            +community.mozillaLink
        engineManager.dtd
            ERROR: reference is a CSS spec at line 14, column 37 for engineManager.style
        pageInfo.dtd
            WARNING: reference is a number at line 55, column 33 for pageInfoWindow.height
            WARNING: reference is a number at line 54, column 33 for pageInfoWindow.width
        places
          moveBookmarks.dtd
              ERROR: reference is a CSS spec at line 2, column 36 for window.style
          places.dtd
              WARNING: reference is a number at line 16, column 32 for places.library.height
              WARNING: reference is a number at line 15, column 32 for places.library.width
        preferences
          advanced-scripts.dtd
              ERROR: reference is a CSS length at line 39, column 41 for window.width
          applicationManager.dtd
              ERROR: reference is a CSS spec at line 2, column 31 for appManager.style
          colors.dtd
              ERROR: reference is a CSS length at line 3, column 43 for window.macWidth
              ERROR: reference is a CSS length at line 2, column 43 for window.width
          connection.dtd
              ERROR: reference is a CSS length at line 16, column 41 for window.macWidth
              ERROR: reference is a CSS length at line 15, column 41 for window.width
          cookies.dtd
              ERROR: reference is a CSS length at line 15, column 45 for window.width
          fonts.dtd
              ERROR: reference is a CSS length at line 15, column 51 for window.macWidth
              ERROR: reference is a CSS length at line 16, column 51 for window.unixWidth
              ERROR: reference is a CSS length at line 14, column 51 for window.width
          languages.dtd
              ERROR: reference is a CSS length at line 1, column 57 for window.width
          permissions.dtd
              ERROR: reference is a CSS length at line 8, column 39 for window.width
          preferences.dtd
              ERROR: reference is a CSS spec at line 21, column 41 for prefWinMinSize.styleGNOME
              ERROR: reference is a CSS spec at line 20, column 41 for prefWinMinSize.styleMac
          preferences.properties
              ERROR: argument 1 `s` should be `S` at line 16, column 16 for phishBeforeText
        safeMode.dtd
            ERROR: reference is a CSS length at line 52, column 39 for window.width
      overrides
        appstrings.properties
            WARNING: trailing argument 1 `S` missing at line 54, column 20 for confirmRepostPrompt
        netError.dtd
            ERROR: not well-formed (invalid token) at line 109, column 61 for cspFrameAncestorBlocked.longDesc
    layout
      css.properties
          WARNING: trailing argument 1 `S` missing at line 64, column 17 for PEAtNSUnexpected
  toolkit/chrome
    global
      customizeToolbar.dtd
          ERROR: reference is a CSS spec at line 2, column 35 for dialog.style
      datetimepicker.dtd
          WARNING: reference is a number at line 3, column 33 for firstdayofweek.default
    mozapps
      downloads
        downloads.dtd
            WARNING: reference is a number at line 17, column 43 for window.height
            WARNING: reference is a number at line 16, column 43 for window.width2
      extensions
        blocklist.dtd
            ERROR: reference is a CSS spec at line 2, column 38 for blocklist.style
        extensions.dtd
            ERROR: reference is a CSS length at line 212, column 47 for eula.height
            ERROR: reference is a CSS length at line 211, column 47 for eula.width
        extensions.properties
            ERROR: argument 2 `S` obsolete at line 114, column 11 for eulaHeader
            +type.dictionary.name
      handling/handling.dtd
          ERROR: reference is a CSS length at line 2, column 26 for window.emHeight
          ERROR: reference is a CSS length at line 1, column 25 for window.emWidth
      xpinstall/xpinstallConfirm.dtd
          ERROR: reference is a CSS spec at line 4, column 34 for dialog.style
my:
errors: 28
missing: 73
keys: 961
unchanged: 998
changed: 4484
missingInFiles: 354
obsolete: 108
warnings: 9

I've left out the missing and obsolete strings, you'll find them elsewhere.

In browser/defines.inc, MOZ_LANGPACK_CONTRIBUTORS should list the real human beings that are the lead contributors, not the dummy names.

In intl.properties,
general.useragent.locale should be my
Is the plural rule the right one?
intl.accept_languages should list my as one of the preferred languages, and maybe other spoken in Myanmar. Arky, can you assist there?
intl.menuitems.insertseparatorbeforeaccesskeys should be true, and not a translation thereof.

Looking forward to the next iteration, and sorry for the lag.
Attachment #569340 - Flags: review?(l10n) → review-
Axel I have no idea how to resolve these compare-locales errors. Can you help me ?
A good start would probably to reset all the CSS specs back to their en-US values, same for the numbers.
I am unable to find this file. Is it obsolete?
xpinstall/xpinstallConfirm.dtd
Attachment #569340 - Attachment is obsolete: true
Attachment #588020 - Flags: review?
Attachment #588020 - Flags: review? → review?(l10n)
xpinstall/xpinstall* moved to extensions/ in toolkit/chrome/mozapps.
Thanks Axel, Please kindly review the code. :D
Status: UNCONFIRMED → NEW
Ever confirmed: true
Comment on attachment 588020 [details]
Compare-Locales Errors Fix, Request Code Review

r- on this one again, sorry.

There are still compare-locales errors. Actually, I even had to update compare-locales to not crash on this.

Also, the previous review comments haven't been addressed yet.

In browser/defines.inc, MOZ_LANGPACK_CONTRIBUTORS still references dummy accounts, while the MOZ_LANGPACK_CREATOR was modified. That second one could have well stayed to be mozilla.org, as in the end, the ones from hg come from mozilla.

intl.properties doesn't seem to be fixed at all.

Here's what compare-locales 0.9.4 shows for this attachment:

my
  browser
    README.txt
        // remove this file
    chrome/browser
      aboutDialog.dtd
          ERROR: not well-formed (invalid token) at line 18, column 37 for community.mozillaLink
          WARNING: Referencing unknown entity `မော်ဇီလာကွန်မြူနတီ` (vendorShortName known) at line 18, column 0 for community.mozillaLink
      devtools/webconsole.properties
          WARNING: trailing argument 1 `S` missing at line 121, column 49 for NetworkPanel.responseBodyUnableToDisplay.content
      inspector.properties
          // remove this file
      preferences/preferences.properties
          ERROR: argument 1 `s` should be `S` at line 6, column 16 for phishBeforeText
      scratchpad.dtd
          // remove this file
      scratchpad.properties
          // remove this file
      styleinspector.dtd
          // remove this file
      styleinspector.properties
          // remove this file
  toolkit/chrome
    global/webConsole.dtd
        // remove this file
    mozapps
      extensions/selectAddons.dtd
          ERROR: reference is a CSS spec at line 17, column 38 for select.action.style
          ERROR: reference is a CSS spec at line 13, column 38 for select.keep.style
          ERROR: reference is a CSS spec at line 23, column 38 for select.name.style
          ERROR: reference is a CSS spec at line 1, column 38 for upgrade.style
      profile/createProfileWizard.dtd
          ERROR: reference is a CSS spec at line 2, column 39 for window.size
      xpinstall
        xpinstallConfirm.dtd
            // remove this file
        xpinstallConfirm.properties
            // remove this file
Attachment #588020 - Flags: review?(l10n) → review-
Thanks Axel, Not sure about the accept languages for Burmese locale (Please help with intl.properties file)
Attachment #588020 - Attachment is obsolete: true
Attachment #590407 - Flags: review?(l10n)
(In reply to arky from comment #16)
> Thanks Axel, Not sure about the accept languages for Burmese locale (Please
> help with intl.properties file)

There's no clear delineation of languages, as non-Burmese languages in Myanmar are somewhat scattered.

However, based on linguistic relationship and population of speakers in Myanmar, I propose this value for intl.accept_languages for Burmese [my]:
my, ybd, rki, kac, ksw, kjp, shn, rhg, prk
(Burmese, Yangbye, Rakhine, Jingpho, S'gaw Karen, Eastern Pwo Karen, Shan, Rohingya, Parauk Wa)

If you have more information about what other languages speakers of Burmese are likely to know, that could help in determining whether this value is useful and appropriate. Also, I have not seen any evidence to suggest that having English (of any form) as a fallback would be useful for speakers of Burmese.
(In reply to Gordon P. Hemsley [:gphemsley] from comment #18)
> If you have more information about what other languages speakers of Burmese
> are likely to know, that could help in determining whether this value is
> useful and appropriate. Also, I have not seen any evidence to suggest that
> having English (of any form) as a fallback would be useful for speakers of
> Burmese.

Oh, I missed this note on Ethnologue: "Native Burmese speakers seldom speak a second indigenous language. If they have one, it is usually English."

As such, maybe add the generic English to the end:
my, ybd, rki, kac, ksw, kjp, shn, rhg, prk, en

Arky, what are your thoughts on the matter?
Gordon, I have to agree with the Ethnologue note. Adding English as fall back might be a good idea. 

Waiting for further feedback from My L10N team.
(In reply to Gordon P. Hemsley [:gphemsley] from comment #19)
> As such, maybe add the generic English to the end:
> my, ybd, rki, kac, ksw, kjp, shn, rhg, prk, en
> 
> Arky, what are your thoughts on the matter?

Last comment of the night. (I shouldn't comment when I'm tired.)

It seems Jingpho and Rohingya are written with the Latin script, while the others are written with the Burmese script (or possibly are unwritten). If this is undesirable (though I'm not sure it is, given potential knowledge of English), I propose the following alternative:
my, ybd, rki, ksw, kjp, shn, prk, en
(In reply to Gordon P. Hemsley [:gphemsley] from comment #19)
> (In reply to Gordon P. Hemsley [:gphemsley] from comment #18)
> > If you have more information about what other languages speakers of Burmese
> > are likely to know, that could help in determining whether this value is
> > useful and appropriate. Also, I have not seen any evidence to suggest that
> > having English (of any form) as a fallback would be useful for speakers of
> > Burmese.
> 
> Oh, I missed this note on Ethnologue: "Native Burmese speakers seldom speak
> a second indigenous language. If they have one, it is usually English."
> 
True. We communicate English with ethnics & neighboring countries except China.
> As such, maybe add the generic English to the end:
> my, ybd, rki, kac, ksw, kjp, shn, rhg, prk, en
> 
> Arky, what are your thoughts on the matter?
Please find Languages in Myanmar with http://dl.dropbox.com/u/1919391/language_tags.pdf Those languages are widely used in Myanmar and assigned in BCP417. Btw, you put rhg (rohingya) in the list, it happens, you must try other languages from neighboring countries. I want to exclude from it. It should be filed in Bangladesh. And some are not languages, It is dialect. ybd (Yangbye).

HTH. --NT
Languages in Myanmar with name in native, name in english and abbreviation.
Comment on attachment 590407 [details]
Checked with compare-locales

Pike, Please kindly do the code review for this bug.
Comment on attachment 590407 [details]
Checked with compare-locales

r=me, there are a few nits left that I fixed with Arky.

There's a bit of confusion of what's currently exported from narro, probably want to fix that by doing a back-import from hg once we have that repo.
Attachment #590407 - Flags: review?(l10n) → review+
fwiw, the accept-language with just my and en-US, en is what we're going with. 80/20 for the win, it's not about covering languages people might possibly speak, but that the majority is likely to be able to read.

We're going to track the further progress towards shipping the Burmese version in the bugs blocking the release tracker bug 723958, and bugs in the Burmese bugzilla component in general (we have a bug to create that).

Welcome to our localization teams, and congrats for the effort up to here.
Status: NEW → RESOLVED
Closed: 8 years ago
Resolution: --- → FIXED
Pike, Please let me know if you need any further feedback on this.
Arky if we should trace back all the bugs of our projects my-mm. This is what I found the earliest.
Let me clear if any wrong like spaces and non-unicode problem,those are what we must solve it up. But I think the last source we can get updated most is from narro and I don't know how to request the extract from narro and how to standardize the locale.

Thanks you guys !
I fixed some string like in /browser/defines.inc.I added some most outstanding contributor names there.

<em:contributor>တင်အောင်လင်း</em:contributor> <em:contributor>ဝေဖြိုးဟန်</em:contributor><em:contributor>ချစ်သီရိမောင်</em:contributor><em:contributor>ရက်ဆဲကျော်</em:contributor><em:contributor>ကျော်စွာသန့်</em:contributor><em:contributor>စည်သူသွင်</em:contributor>
You need to log in before you can comment on or make changes to this bug.