Re: What's cooking in git.git (Sep 2010, #01; Wed, 1)

From: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Date: Sep 1, 2010, 20:22 UTC
Message-ID: <AANLkTim6=OTwzRd2H0GzNO3T0JuJxox8-Hfs_iKV=Xxy@mail.gmail.com>
In-Reply-To: <7vaao15jk2.fsf@alter.siamese.dyndns.org>

On Wed, Sep 1, 2010 at 19:51, Junio C Hamano <gitster@pobox.com> wrote:

Show 7 quoted lines

> As I didn't want to worry about the early parts of ab/i18n topic, which
> was accidentally merged to 'next' due to my lazyness (another topic,
> ab/test, seemed to be ready for the upcoming release, but a patch to that
> topic had interaction with ab/i18n, which made ab/test unmergeable to
> 'master' without dragging 'ab/i18n'), I reverted ab/i18n topic from 'next'
> and merged a rebuilt ab/test topic to 'next'.  The i18n topic is merged
> to 'pu'.

Cool, whatever makes release engineering easy for you. But the "gettext: Make NO_GETTEXT=YesPlease the default in releases" monkeypatch should probably be ejected out of pu.

I only wrote it in case gettext wouldn't be ejected from next and would make it into 1.7.3, but since that won't be the case there's really no reason for it to exist anywhere but the mailing list archive.

It's also broken as a proper way to disable gettext (in case we want to go that route). There's no way to re-enable it if you use the ./configure script.

It's really only a monkeypatch.

Show 5 quoted lines

> * ab/test-2 (2010-08-30) 51 commits
>  (this branch is used by ab/test.)
>
> As I'd like to have this in the upcoming release, the tip of old ab/test
> topic has been rebuilt without the merge from ab/i18n topic.

Thanks, and sorry for tangling them up.

Show 6 quoted lines

> * kf/askpass-config (2010-08-30) 3 commits
>  - Extend documentation of core.askpass and GIT_ASKPASS.
>  - Allow core.askpass to override SSH_ASKPASS.
>  - Add a new option 'core.askpass'.
>
> Should be Ok to merge to 'next'; then will merge to 'master' shortly.

Is this facility also used for the smtp password promt in git-send-email? Should it be?

Show 10 quoted lines

> * jj/icase-directory (2010-08-16) 6 commits
>  (merged to 'next' on 2010-08-26 at 9d8e1bc)
>  + Support case folding in git fast-import when core.ignorecase=true
>  + Support case folding for git add when core.ignorecase=true
>  + Add case insensitivity support when using git ls-files
>  + Add case insensitivity support for directories when using git status
>  + Case insensitivity support for .gitignore via core.ignorecase
>  + Add string comparison functions that respect the ignore_case variable.
>
> Depends on GNU FNM_CASEFOLD.

We're going to have to get this working on non-GNU if it's to make it into 1.7.3. I can probably help with that if jj doesn't have time.

Show 15 quoted lines

> * ab/compat-regex (2010-08-26) 9 commits
>  (merged to 'next' on 2010-08-26 at c5cc9d2)
>  + Fix compat/regex ANSIfication on MinGW
>  (merged to 'next' on 2010-08-22 at 40bce7c)
>  + autoconf: regex library detection typofix
>  (merged to 'next' on 2010-08-21 at 632d60f)
>  + autoconf: don't use platform regex if it lacks REG_STARTEND
>  + t/t7008-grep-binary.sh: un-TODO a test that needs REG_STARTEND
>  + compat/regex: get rid of old-style definition
>  + compat/regex: define out variables only used under RE_ENABLE_I18N
>  + Change regerror() declaration from K&R style to ANSI C (C89)
>  + compat/regex: get the gawk regex engine to compile within git
>  + compat/regex: use the regex engine from gawk for compat
>
> Will merge to 'master' shortly.

I'll try to recall to get these patches we have on top of it into gawk/glibc once it's settled a bit more.

Show 5 quoted lines

> * ab/i18n (2010-08-31) 23 commits
> [...]
> Kicked back to 'pu' to give it a bit more time to experiment with a
> Porcelain command or a two, and an opportunity for a fresh restart if
> necessary in the next cycle.

Incidentally I'm just about to submit some more experiments :)