Gitlab experiment status

[Rémi Bernon]

On 5/25/22 20:56, Alexandre Julliard wrote:
> Rémi Bernon <rbernon at codeweavers.com> writes:
> 
>> More precisely, the rough edges are a bit related to the current patch
>> submission process, and I think they could be smoothed out if Gitlab
>> was the preferred development platform:
>>
>> * The sign-off requirement makes it a bit more painful to create MR
>>    than it could be. There's git rebase --sign-off, but it's a bit
>>   inconvenient to script. As other have mentioned already, reviewer
>> information is also not included in commits anymore, which feels like
>> a loss.
> 
> The current idea is to use git notes to propagate that information from
> Gitlab, to avoid having to rewrite MR commits. There's an experimental
> refs/notes/commits branch on the main gitlab tree that you can pull if
> you'd like to see what it looks like. If this works out we could
> potentially use it for the original author signoff as well.
> 


That's neat, I didn't know it was there already!

FWIW it can be configured to be fetched automatically into the local 
notes with:

   git config --add remote.origin.fetch "refs/notes/*:refs/notes/*"

Then `git log` will automatically show notes for each commit that went 
through Gitlab.

Though beware that if you're already using `git notes` locally, this 
will cause some conflicts and you need to decide which notes branch to use.

And if you don't use git notes locally, you can safely a '+' in front of 
the fetch refspec to make sure git fetch will force update it from origin.


Cheers,
-- 
Rémi Bernon <rbernon at codeweavers.com>