On Tue, Nov 13, 2007 at 12:32:19 +0100, Thomas Neumann wrote:
quoted text
> > Is slightly troubles me that you put so much emphasis on what I would c=
all=20
quoted text
> > "remote information".  I understand that in svn, your working directory=
=20
quoted text
> > without the server is not very useful.  But we do not have that problem.
> that is true. My usage pattern probably stems from the fact that I am a
> long term svn user :) And I use git for work now, where there is indeed
> some kind of central repository just as in a Subversion setting.
> In a fully decentralized setting the remote information is probably not
> as important, although you might still want to know what happens if you
> issue "git pull".

The remote URL isn't /the/ useful bit, most of the time. Either you have ju=
st
one remote, which is the project central repository and you probably know
which it is just by knowing which project it is, or you have many of them a=
nd
their names tell you enough.

Note, that unlike in Subversion, the branch name is /not/ part of the URL.
And that is the useful bit of the information. So what 'git info' probably
should show is:
 - Which branch is currently checked out
 - Which branch it is tracking (inspect the config)
 - List of n (where n is small integer) "closest" branches, where the
   distance to a branch is number of commits in HEAD since common ancestor
   with that branch.
 - Latest included tag. Basically something like git describe.
 - Short log of last few commits.

--=20
						 Jan 'Bulb' Hudec <bulb@ucw.cz>