l10n: the maintaince of git-po repo

I'm looking forward to the next git release with Chinese l10n, so that I can start to rewrite my book on Git. ;-)

The maintaince of the git-po and git-po-zh_CN repositories so far:

git-po repository
-----------------
Location   : https://github.com/gotgit/git-po
Fork from  : https://github.com/git/git
Description: This is the Git l10n master repo, and is used to coordinate
             works of l10n teams. L10n team leaders send pull requests
             to this repo, while other l10n contributors should wok on
             the language specified repository created by the l10n team
             leader.
Typical commits:

- commit https://github.com/gotgit/git-po/commit/71eb878

      merge the pull request sent from the Chinese l10n team leader.

- commit https://github.com/gotgit/git-po/commit/1d4ac88

      l10n: Update git.pot (12 new messages)

- commit https://github.com/gotgit/git-po/commit/508d124

      Merge branch 'master' of 'git.git' into git-po,
      and will generate new po/git.pot in next commit.

- commit https://github.com/gotgit/git-po/commit/dce37b6

      l10n: initial git.pot for 1.7.10 upcoming release

git-po-zh_CN repository
-----------------------
Location   : https://github.com/jiangxin/git-po-zh_CN
Fork from  : https://github.com/gotgit/git-po
Description: Git l10n for Simplified Chinese.
Typical commits:

- commit https://github.com/jiangxin/git-po-zh_CN/commit/7ac1c0a

      l10n: update Chinese translation to the new git.po

- commit https://github.com/jiangxin/git-po-zh_CN/commit/0ad9e96

      fix of Chinese translation

- commit https://github.com/jiangxin/git-po-zh_CN/commit/d4a8e00

      fix of Chinese translation

- commit https://github.com/jiangxin/git-po-zh_CN/commit/c04f5ac

      l10n: initial of zh_CN l10n

Jiang Xin <worldhello.net@gmail.com> writes:

For people who don't know, Jiang kindly sent me an autographed copy of his book https://plus.google.com/u/0/108182106411180467879/posts/XX2CewgxSAb I can only read the command examples, but the book seems to have a good coverage of modern Git, judging from the table of contents.

Thanks.  I took a look.

I'll comment on the changes as if they came in the patch form here, to ask for sanity checks from the list members.

The file is no longer a generated file but is tracked, and should not be removed upon distclean. Good.

Likewise, it is not ignored anymore. Good.

Should the above placeholder lines left as they are, or should they be updated to name "Git" as the "PACKAGE" with "SOME DESCRIPTIVE TITLE"?

If the answer is "latter", I think we would need to apply a similar update to po/git-gui.pot for git-gui package.

On the other hand, I think the placeholder lines below should stay as they are.

Describe the high-level overview of the localization workflow, and update the description on po/git.pot file, which no longer is an untracked file.

Clarify the roles and responsibilities among:

 - the core developers (responsible for marking the translatable strings
   and restructuring the code to avoid message-lego);

 - the l10n coordinator (maintaining po/git.pot, interfacing with language
   teams and feeding updated translations back to the core); and

 - the language teams (updating po/*.po for their language and feeding
   their work back to the l10n coordinator).

This is largely based on an earlier message $gmane/189584 from me.

  Date: Wed, 01 Feb 2012 20:56:18 -0800
  Message-ID: <7vmx91ygst.fsf@alter.siamese.dyndns.org>

Signed-off-by: Junio C Hamano <gitster@pobox.com>
---

 * I would like to see something like this before we start getting the
   second language team.  Comments are very much appreciated.

   I am not sure if the existing instruction to update the new XX.po
   file copied from git.pot is correct, but I left it as-is.

    @@ -1,6 +1,6 @@
    -# Icelandic translations for PACKAGE package.
    -# Copyright (C) 2010 THE PACKAGE'S COPYRIGHT HOLDER
    -# This file is distributed under the same license as the PACKAGE package.
    +# Icelandic translations for Git.
    +# Copyright (C) 2010 Ævar Arnfjörð Bjarmason <avarab@gmail.com>
    +# This file is distributed under the same license as the Git package.
     # Ævar Arnfjörð Bjarmason <avarab@gmail.com>, 2010.

   The second line of the template tells us to name the copyright holder
   of the package, not of the translation, but the above example makes it
   appear as if we are assigning the copyright to the translator.

 po/README |  106 ++++++++++++++++++++++++++++++++++++++++++++++++-------------
 po/TEAMS  |   10 ++++++
 2 files changed, 94 insertions(+), 22 deletions(-)
 create mode 100644 po/TEAMS

2012/2/29 Junio C Hamano <gitster@pobox.com>:

I create a new dedicated team, named git-l10n, and transfer this repo to the new place: https://github.com/git-l10n/git-po/ for the following reasons:

 - git-po is a long-term repo, so has a dedicated, stabilized URL is better.
   The repo URL may not change if l10n coordinator change.

 - git-l10n is a github organization account, and has a dedicated owners
   group, who can assign new l10n coordinator.

 - Owners group of git-l10n organization will not receive pull requests to
   git-po. I add gister, avar to the owners team.

 - Members of the git-l10n-coordinators team of git-l10n organization
   will receive pull requests.

 - Some Chinese guys used to send pull request to gotgit/git-po, so you
   can find some closed pull requests written in Chinese. It won't happen
   after the repo transfer to git-l10n/git-po.

As a contributor for a language XX, you should first check TEAMS file in this directory to see whether a dedicated repository for your language XX exists. Fork the dedicated repository and start to work if it exists.

If you are the first contributor for the language XX, please fork this repository,

Shall we remind the contributors for l10n to write commit log in English, but not language XX? It's a common mistake for translators.

I think we should add the URL of the dedicated repository for language XX here. Example:

Language:      zh_CN (Simplified Chinese)
Leader:           Jiang Xin <worldhello.net@gmail.com>
Homepage:     https://github.com/jiangxin/git-po-zh_CN/

Contributors and members can add to the header of the XX.po file. Add here may cause the TEAMS file grows rapidly.

Jiang Xin <worldhello.net@gmail.com> writes:

It is unclear the distinction between gotgit and git-l10n but I am guessing that (1) you feel "gotgit" is more for _your_ personal use, (2) you happen to be the volunteer coordinator right now, but (3) the team effort deserves more permanent home, detached from your identity.

And if that is the case, I agree 100% with the reasoning.

That's better.

I would prefer to see the subject line (the first line summary) in English to say something like "Update zh_CN translation for 1.7.12". That much of English shouldn't be too much burden for language teams.

On the other hand, the body of the log that describes the change _may_ be hurt if you forbid the use of language XX. The translators may want to write something like:

	The earlier translation used word '[%#$@$#]' for original phrase
	'to distim' in English, but word '[**&*#$]' is a more appropriate
        language XX translation in the context the message for such and
	such reasons.

and I suspect that "for such and such reasons" part is far easier to explain and understand if written in language XX for native speakers of language XX. Such a description often has to discuss connotations and subtle nuances of the particular word that is hard to explain.

Good thinking; I don't know if Homepage or Repository is more appropriate, but some URL would definitely help.

Sorry, I do not understand what you are worried about here. Could you clarify a bit?

In any case, the patch you are replying to is "I'd like to see *something* *like* *this*...", so feel free to take the ownership and update it. I'll try to help refining it if you send an updated draft to back the list for review.

Thanks.

2012/3/1 Junio C Hamano <gitster@pobox.com>:

I noticed that all i18n related commit logs all have prefix "i18n:". It would be nice to add a prefix "l10n:" to the l10n related commit log, such as: "l10n: Update zh_CN translation for 1.7.12".

Git does not handle multi-bytes character well. Multi-bytes characters not convert to UTF-8 when git write tree objects and commit objects. I think allow multi-bytes characters in commit log would hurt git.git until git gives full support UTF-8 (are there any plans for this?). I just tested several versions of msysgit:

 - Only Git-1.7.8-preview20111229-unicode.exe saves commit objects
   with utf-8 encode.

 - Other versions such as Git-1.7.9-preview20120201.exe and
   Git-1.7.8-preview20111206.exe will not convert multi-bytes characters
   in commit log into UTF-8.

Am 3/2/2012 8:30, schrieb Jiang Xin:

It does not convert, but it records which encoding the text has. If you don't specify anything, UTF-8 is assumed, and if your text is actually not UTF-8, the result is necessarily garbage.

Then you haven't set i18n.commitEncoding. Try this:

   git config i18n.commitEncoding CP936

(substitute whatever Windows codepage you are using).

-- Hannes

2012/3/2 Johannes Sixt <j.sixt@viscovery.net>:

I know there are two config variables. i18n.commitEncoding will insert a "encoding XX" line to the commit object, while i18n.logOutputEncoding will set the default output encoding. But this implementation seems like a workaround.

* Tree objects do not have such implementation, so multibyte characters
   can not be used as filenames.

* Commit object without "encoding" instruction will be used as it is. So
  people under the same non-utf8 locale may not notice that they
  have not set the proper i18n.commitEncoding, until one day they
  need accross platform development.

* As raw commit log alway used as fallback, sometimes it is hard to find
  commit objects with wrong encoding instructions.

I think save commit object, tree object, packed-refs in UTF-8 is a better implementation.

Jiang Xin <worldhello.net@gmail.com> writes:

Note that according to documentation 'git commit' issues a warning if the commit log message given to it does not look like a valid UTF-8 string, unless you explicitly say your project uses a legacy encoding. Modern git would also warn if you have NUL ("\0") character in your commit message, e.g. when using UCS-2 / UTF-16 encoding.

And there is no place on pathnameEncoding in 'tree' object, unfortunately.

One proposed solution was to convert filenames from filesystem encoding to normal-form composed UTF-8 when creating tree objects, but this would have to be optional.

Anyway at least for source code using characters outside US-ASCII is
really discouraged anyway.  Note also that sample 'pre-commit' hook
prevents adding files with non-ascii filenames.
 

But see above.
 

Backward compatibiltiyt and performance.

Jiang Xin <worldhello.net@gmail.com> writes:

Even outside i18n/l10n context, projects that want to be usable by wide population avoid filesystem-unsafe characters, which are not limited to just multi-byte. Even though Git started on Linux, we do not have a file whose name contains '\' or HT, which would have made it impossible to check out on DOS.

It's just "common sense" or "common courtesy".

In any case, I do not expect we would need to have po/國語.po in git-l10n repository, so this is a non issue in the context of this discussion thread.

That is why we have l10n coordinator who can vet such mistakes crawling into our system, isn't it?

It is one of the better conventions to encourage.

It is entirely a different matter to forbid a closed local user community from adopting GB, Big5, EUC or whatever character encoding as the encoding of their choice, and using it throughout their project for pathnames, blob payloads and log messages.

Host the l10n coordinator repository in a dedicated github organization account "git-l10n", so that the team may have a more permanent home. Also add a hint about reference of TEAMS file for l10n contributors.

Update TEAMS file with new zh_CN l10n team members and a repository URL.

Signed-off-by: Jiang Xin <worldhello.net@gmail.com>
---
 po/README |   17 +++++++++--------
 po/TEAMS  |    9 +++++++--
 2 files changed, 16 insertions(+), 10 deletions(-)