Re: [PATCH v2 17/20] transport-helper: change import semantics
- From
Jonathan Nieder <jrnieder@gmail.com>
- Date
- Jul 4, 2011, 22:37 UTC
- Message-ID
- <20110704223759.GC8909@elie>
- In-Reply-To
- <CAGdFq_hFQRTjRZBGZKBejpUqCCzo+wHWG69nM8=57peUBN9yjw@mail.gmail.com>
Sverre Rabbelier wrote:
> On Mon, Jul 4, 2011 at 23:58, Jonathan Nieder <jrnieder@gmail.com> wrote:
>> It would mean no change compared to the current behavior. :)
[...]
> With the only difference that currently that \n signifies the end of > the entire stream, whereas now it will mean the end of the imports?
Exactly. The git-remote-testgit part of the patch would be pretty similar to what you have and could even go earlier in the series if wanted (since it does not require transport-helper changes), while the transport-helper part should be unnecessary as long as patch 15/20 "use the new done feature where possible" does not remove the \n in its output (as discussed in its subthread).
Meanwhile I should probably implement a "capability done" as Dmitry hinted, to make blank lines no longer act as a terminator as a way of helping people debugging by hand that hit <ENTER> too many times by mistake.