Re: git-cl under Windows

[Trevor Daniels]

David Kastrup wrote Tuesday, July 03, 2012 10:46 PM

> Graham Percival <address@hidden> writes:
> 
>> On Tue, Jul 03, 2012 at 10:26:18PM +0200, David Kastrup wrote:
>>> "Trevor Daniels" <address@hidden> writes:
>>> 
>>> >> Description empty; aborting.
>>> 
>>> That just means that git-cl called what it considered an editor, and the
>>> file for editing did not change, so git-cl aborted.
>>
>> Another option would be to change git-cl to skip this check -- I
>> mean, you almost always want the rietveld message to just be the
>> git commit message anyway.
> 
> Can't agree: when I put up a review message, I often put in additional
> info/rationale that may be relevant for reviewers, but not really
> belongs in a commit message.

I've now got git-cl working using Python 2.7.  This is definitely the
easier route under Windows.  At present it just takes the commit
message, as I'd already patched around the call to vi before
I saw David's suggestion, but I should be able to put that back now
I have a working version.  For now I can easily add info to reviewers
in the first comment.  You have to create this to announce the new
patch-set anyway.

Just need to do some tidying up now after all my hacking around.

Thanks both for your help.

Trevor