Bzr shouldn't permit users to commit without setting author info

[Martin Albisetti]

Bzr shouldn't allow users to commit if they don't have any author info specified, and explain to them how they should set it.
As a bonus, I'd like to add with what author info you are committing on each commit :)

[John A Meinel (jameinel)]

Not sure what the "bonus" means.

Are you saying you shouldn't be allowed to commit without setting "bzr whoami"? Or is this more about the --author flag?

[Martin Albisetti (beuno)]

By "bonus" I meant it's something I'd like to add too (and I'm current working on a patch for it).
Might be a separate thing, but it addresses the same issue: try to avoid users committing with wrong "whoami" information.

Yes, by author I do mean "whoami".
What would be the correct term for this?

[John A Meinel (jameinel)]

Technically, that is the 'committer'.

At the moment, we let you commit, but we guess the committer based on your username and the hostname of the machine you are on. We've talked about removing the automatic detection. I think the idea was that having only 3 steps to get going "init; add; commit" was nice enough that we were trying to avoid an extra "bzr whoami XXXX" step. It isn't strictly required, it only depends on how much people care to have exactly the right info in the commit messages.

[Martin Albisetti (beuno)]

While I understand this would introduce an extra step, it seems we are running more and more into users wanting to change their previous commits because they didn't set the committer.

And, since bzr can't modify previous revisions, than I think the tradeoff is worth it.

Of course, if the majority thinks otherwise, I'll refrain from working in this any longer :)

(although, reporting what the committer info was on each commit is still useful, and doesn't change anything the current workflow)

[John A Meinel (jameinel)]

I probably prefer refusing to commit, as I've switched machines and then you see some "jameinel@XXXX" commits which I didn't intend.

I was just letting you know that this has been discussed in the past.

[Martin Albisetti (beuno)]

This should still be done, but I will likely not get to it.