Re: [Monotone-devel] bugfest analysis

[Timothy Brownawell]

On 05/10/2010 03:44 PM, Thomas Keller wrote:
> Am 10.05.10 05:23, schrieb Timothy Brownawell:
> > On 05/09/2010 06:28 PM, Thomas Keller wrote:
> >
> > > * #17878: Usability: too easy to accidentally fork after
> > >     merge or disapprove
> > >     ->   branch is net.venge.monotone.bugfest-2010.17878-tbrownaw
> > >     ->   @Tim: is this done?
> >
> > It needs a test using the --update option, which I'll get to either in
> > the morning or after work tomorrow. I suppose I should also put a note
> > in NEWS about setting your get_default_command_options() if you want to
> > always have that behavior.
>
> Looks cool so far, a few minor questions / nitpicks:
>
> 1) In maybe_workspace_updater.cc:
> +    if (rev.edges.size() != 1)
> +      return not_updatable;
>
> Wouldn't that mean that the root revision of a branch is not updatable?
> I.e. create a new project, commit and push that. Somebody else changes
> it, you pull with --update afterwards, or do I get something wrong?

That still has one edge, from a parent with the null revision id.

> 2) "Your workspaces have not been updated" - why are you using plural here?

Hm. I guess because I tend to have multiple workspaces (because I'm 
fairly easily distractable and don't work in small enough chunks yet). 
But to really be consistent with this it should say "your other 
workspaces have not been updated" when it *does* update.

...I suppose the having too many workspaces really is a bad habit we 
don't want to encourage, so I'll go make it not plural. And then it 
doesn't have to say anything when it does update.

--
Timothy

Free public monotone hosting: http://mtn-host.prjek.net