"Registered branches page" doesn't list any branches

[wouter bolsterlee]

https://launchpad.net/people/uws/+registeredbranches doesn't list the branches I registered. They do show up at https://launchpad.net/people/uws/+authoredbranches fine.

[David Allouche (ddaa)]

That is the intended behaviour.

Branches are associated to a registrant and an author, that makes it possible to register branches on behalf another user, or register a branch and later set its author after that person has created a Launchpad account.

That is a ugly hack, the use cases would be better served by a workflow to transfer ownership of objects. See bug 29863.

I am not sure what should happen to the "Registered branches" page and the branch auhor property in the meantime. Maybe they are just confusing and should be removed.

[wouter bolsterlee (wbolster)]

I *am* the registrant of the branch, so it should definitely be listed when listing the branches registered by me.

[David Allouche (ddaa)]

After a chat in #bzr, I was convinced that author is actually useful and that this is actuall an UI issue in personal branch listings.

Here is the real-world use case for author.
<LarstiQ> ok, let me walk through a small use case then
<LarstiQ> say, we have the product bitlbee, with main author wilmer
<LarstiQ> wilmer isn't really interested in launchpad himself, but the project is using bzr on insistance of the other devs
<LarstiQ> now, I'd like to attribute his branches to him
<LarstiQ> waiting for him to login and accept is painful
<ddaa> that's one of the two uses of author != registrant in launchpad: https://launchpad.net/people/jelmer/+branch/bitlbee/main
<LarstiQ> jelmer would be one of the lp inclined devs :)

The UI could be fixed by making the "registered branches" page list all the branches that person is the registrant of, or change the informative text on that page to read "These are branches you registered in Launchpad, but which you didn't author".

I'm not sure which one would be best.

[wouter bolsterlee (wbolster)]

For completeness, another piece of IRC log:

17:00:09 < ddaa> I understand you are the registrant.
17:00:22 < uws> yeah.
17:00:27 < ddaa> But you are also the author.
17:00:31 < uws> Correct.
17:00:41 < ddaa> The way I designed it (I think it's wrong, but that's a separate issue)
17:00:55 < uws> So the affected branches for which I am *BOTH* author and registrant should show up on both pages
17:01:04 < ddaa> When somebody is author _and_ registrant, the "registrant" status is essentially ignored to avoid redundancy.
17:01:16 < ddaa> I understand what you mean.
17:01:47 < uws> Alternatively, the description should read "These are branches you registered in Launchpad, but which you didn't author"
17:01:58 < ddaa> mhh
17:02:16 < uws> ddaa: I understand your point and I agree with it. But from a user perspective it's really confusing
17:02:24 < ddaa> I agree that is confusing.
17:02:47 < ddaa> I'm just pondering whether Branch.author should be just considered "a bad idea" and be removed altoghether.
17:03:44 * ddaa checks the database
17:03:59 < uws> ddaa: No, author is great!
17:04:19 < uws> ddaa: "What does this person hack on?"

In my opinion, changing the descriptive text on the registeredbranches page would be the best option, so that branch information is not duplicated on different pages.

[David Allouche (ddaa)]

heed thy users!

[David Allouche (ddaa)]

Committed a patch to clarify the registeredbranches page. Should be rolled out monday.