[PATCH STABLE] histedit: save manually edited commit message into ".hg/last-message.txt"

[Pierre-Yves David]

On 04/14/2014 01:07 PM, FUJIWARA Katsunori wrote:
>
> At Mon, 14 Apr 2014 01:34:35 -0400,
> Pierre-Yves David wrote:
>>
>> On 04/02/2014 08:17 AM, FUJIWARA Katsunori wrote:
>>>
>>> At Tue, 01 Apr 2014 17:07:42 -0500,
>>> Matt Mackall wrote:
>>>>
>>>> On Mon, 2014-03-24 at 01:01 +0900, FUJIWARA Katsunori wrote:
>>>>> # HG changeset patch
>>>>> # User FUJIWARA Katsunori <foozy at lares.dti.ne.jp>
>>>>> # Date 1395590238 -32400
>>>>> #      Mon Mar 24 00:57:18 2014 +0900
>>>>> # Branch stable
>>>>> # Node ID dd35a6ff046d4516384e32f55741b474b35f72b6
>>>>> # Parent  e5641536e4d509b2dc5fab783344f86ea61b17c2
>>>>> histedit: save manually edited commit message into ".hg/last-message.txt"
>>>>
>>>> Oops, I see I already queued a later version of this (that magically
>>>> managed not to conflict). Dropped.
>>>
>>> If you mean 5d22cadd1938 as "a later version", this patch and
>>> 5d22cadd1938 are different each other.
>>>
>>>       http://selenic.com/repo/hg/rev/5d22cadd1938
>>>
>>> 5d22cadd1938 fixes the problem in "fold" of histedit-ing, and this
>>> patch fixes one in "message".
>>>
>>> I noticed my overlooking this problem after your queuing my series
>>> including 5d22cadd1938.
>>>
>>> Sorry for ambiguous summary line, but 80 characters are not enough for
>>> me to describe not only purpose of patch but also difference between
>>> them, :-)
>>
>> I've re-clowncopterized this for stable. Consider sending a proper patch
>> in default before the end of the week.
>>
>> Thanks for the patch! This family of bug is usually very frustrating.
>
> Thank you for your comments.
>
> And please let me confirm some points.
>
>    (1) is this patch already queued ?

yes it is available there: http://42.netv6.net/clowncopter/

>
>      According to Augie's post, "being clowncopterized" means "being
>      queued".
>
>        http://selenic.com/pipermail/mercurial-devel/2014-April/057954.html
>
>      But you also said "consider sending a proper patch ....". Does
>      this mean that this patch is not yet queued (and you recommend me
>      to resend) ?

You description say: this is a simple fix for stable. I've better idea 
for default. I would be happy to review this idea.

>
>    (2) can it be canceled ?
>
>      I forgot that this patch includes problem fixed by Sean's
>      e259d4c462b5 (using TESTDIR instead of TESTTMP incorrectly) at
>      this replying.
>
>      If it can be canceled, I'll resend revised version of
>      this. Otherwise, I'll send new one fixing this problem after
>      appearance of this patch in the official public repositories.

It did not made it into main yet, it can be rewritten or dropped.

>
>    (3) is this patch suitable for "default" (as you recommend to
>        resend) ?

not sures this matters since the freeze is in less than 100 hours

-- 
Pierre-Yves David