Stamps
Matt Mackall
mpm at selenic.com
Mon May 9 13:08:22 CDT 2011
On Mon, 2011-05-09 at 13:57 -0400, Steve Losh wrote:
> On 05/09/11 at 12:48P, Kevin Bullock wrote:
> > On May 9, 2011, at 12:35 PM, Kevin Bullock wrote:
> >
> > > On May 7, 2011, at 6:52 PM, Adrian Buehlmann wrote:
> > >
> > >> veracity [1] has the concept of "stamps". A stamp is like a tag (a name)
> > >> that can be applied to changesets. But a stamp differs from a tag
> > >> because the same stamp can be applied to *multiple* changesets.
> > >>
> > >> Example usage: "passes test", "behavior change", "pending review"
> > >
> > > How is this different from (a) named branches, and (b) bookmarks when used as described in issue2788?
> > >
> > > For reference, here's the 4(!) types of labels we currently have (cribbed from mpm's notes on issue2788):
> >
> > Actually let's extend that with a column that indicates whether a label can apply to multiple commits or not:
> >
> > name stored mutable auditable sharable auto-updating multiple
> > tags .hgtags yes yes yes no no
> > localtags .hg yes no no no no
> > branches commit no yes yes no yes
> > bookmarks .hg yes no yes if-current no
> >
> > ...the extra column bringing the number of possible combinations to *20*.
>
> Correct, and stamps would fit in like so:
>
> name stored mutable auditable sharable auto-updating multiple
> tags .hgtags yes yes yes no no
> localtags .hg yes no no no no
> branches commit no yes yes no yes
> bookmarks .hg yes no yes if-current no
> stamps ? yes ? yes no yes
The point remains: we're not going to add a dozen different types of
markers to cover all useful behaviors (the above table suggests 64
possible types!). Even having one for each column above would be too
many. Even the four we currently have: too many. I'm planning
deprecating[1] localtags soon.
[1] Before anyone freaks out, that just means it will cease to appear in
the help for tag.
--
Mathematics is the supreme nostalgia of our time.
More information about the Mercurial-devel
mailing list