Re: What's cooking in git.git (Sep 2011, #06; Wed, 21)

From: Junio C Hamano <gitster@pobox.com>
Date: Sep 22, 2011, 16:39 UTC
Message-ID: <7v62kkzg3s.fsf@alter.siamese.dyndns.org>
In-Reply-To: <4E7AFC6C.7080603@elegosoft.com>

Michael Schubert <mschub@elegosoft.com> writes:

Show 10 quoted lines

> On 09/22/2011 10:37 AM, Carlos Martín Nieto wrote:
>> On Wed, 2011-09-21 at 22:04 -0700, Junio C Hamano wrote:
>>> * cn/eradicate-working-copy (2011-09-21) 2 commits
>>>  - patch-id.c: use strbuf instead of a fixed buffer
>>>  - Remove 'working copy' from the documentation and C code
>> 
>> It looks like that first commit sneaked in there. Shouldn't that be its
>> own topic?
>
> It's in pu twice:

Thansk for noticing. Will remove.

Show 12 quoted lines

> On 09/22/2011 07:04 AM, Junio C Hamano wrote: 
>> * cn/eradicate-working-copy (2011-09-21) 2 commits
>>  - patch-id.c: use strbuf instead of a fixed buffer
>>  - Remove 'working copy' from the documentation and C code
>
>> * ms/patch-id-with-overlong-line (2011-09-21) 1 commit
>>  - patch-id.c: use strbuf instead of a fixed buffer
>
> 64128da and a6c5c60
>
> There's also a minor typo in the last sentence of the commit message.
> Should I resend?

If that is s/user/use/ and nothing else, I can amend locally; otherwise please do.

The last part of the "rather dumb heiristic" you talk about is designed to parse "cmd | diff-tree -p --stdin" output. I agree with you that it can safely tightened a bit by checking that 41st byte is the EOL, but it probably is not worth it.