On Tue, Jan 31, 2023 at 01:40:56AM +1100, Mark Jamsek wrote:
> I initially displayed both the
> worktree UUID and ref like so:
> 
> -----------------------------------------------
> cherrypick 862b68b7d40a3aec7cc411c275359a19da6fae73
> from: Mark Jamsek <mark@jamsek.dev>
> date: Thu Jan 26 06:36:17 2023 UTC
> 
>  foo
> 
>  M  node
> 
> work tree: 6a68c142-68ca-4237-8fd3-96de08cc7306
> refs/got/worktree/cherrypick-6a68c142-68ca-4237-8fd3-96de08cc7306-862b68b7d40a3aec7cc411c275359a19da6fae73
> 0
> 
> But I'm not sure the full ref path adds much. It can be surmised by
> appending the type (i.e., "backout" or "cherrypick"), which is now
> displayed in the header, and the UUID and commit hash--both now also in
> the output--to the namespace (i.e., "refs/got/worktree"), which got(1)
> mentions. So all the information is present already. Nonetheless, I'm
> not sure what the full path adds.
> 
> That said, I'm completely fine with displaying it if we think it's
> a good idea!

I agree that displaying the full path is redundant.
Even without it there is enough information being displayed to make
sense of things.
 
> The diff is really simple, but there's a bit of churn in regress to
> account for the header change. I've not amended regress further to test
> for the "work tree: UUID" output till we're settled on the extent of the
> output.

Sure, more testing can be added later.  ok by me.