threads / discuss / 7269

Question about bare repositories

Subject: Question about bare repositories

## tl;dr

4 messages between Mar 16, 2007 and Mar 16, 2007.

replies: 3people: 3as markdown or json

Andy Parkins· Mar 16, 2007, 09:47 UTC · lore

Hello,

I'm having an understanding problem, rather than a git problem. It came about because I was looking at a repository using gitweb. The repository was bare, and happened to have HEAD pointing at an old branch that hadn't been updated for a long time. I was confused because I wasn't seeing new commits. Once I set HEAD to master in the bare repository, gitweb showed the new commits.

All that led me to this question, which isn't a criticism, but has exposed a hole in my understanding of git.

Why does a bare repository have/need a HEAD?

Andy

-- 
Dr Andy Parkins, M Eng (hons), MIET
andyparkins@gmail.com

Junio C Hamano· Mar 16, 2007, 10:05 UTC · re: Andy Parkins · lore

Re: Question about bare repositories

Andy Parkins <andyparkins@gmail.com> writes:

> Why does a bare repository have/need a HEAD?

To mark which branch the repository owner considers the primary branch of the repository. I think "git pull $URL" (i.e. without using shorthand arranged in the config nor $GIT_DIR/remotes/) defaults to fetching and merging HEAD from there, for example (I say "I think" because it's been a while I did so myself, and we might have broken it during the 1.5.0 process, which had a lot of work in that area).

Andy Parkins· Mar 16, 2007, 11:52 UTC · re: Junio C Hamano · lore

Re: Question about bare repositories

On Friday 2007 March 16 10:05, Junio C Hamano wrote:

Show 5 quoted lines

> Andy Parkins <andyparkins@gmail.com> writes:
> > Why does a bare repository have/need a HEAD?
>
> To mark which branch the repository owner considers the primary
> branch of the repository.  I think "git pull $URL" (i.e. without

Interesting. So it's got a much reduced importance and sounds like it is mainly as guidance for cloners, etc.

Show 5 quoted lines

> using shorthand arranged in the config nor $GIT_DIR/remotes/)
> defaults to fetching and merging HEAD from there, for example (I
> say "I think" because it's been a while I did so myself, and we
> might have broken it during the 1.5.0 process, which had a lot
> of work in that area).

Me too; I don't think I've ever used it, which is where my confusion came from.

Thanks for the help.

Andy

-- 
Dr Andy Parkins, M Eng (hons), MIET
andyparkins@gmail.com

Johannes Schindelin· Mar 16, 2007, 12:24 UTC · re: Andy Parkins · lore

Re: Question about bare repositories

Hi,

On Fri, 16 Mar 2007, Andy Parkins wrote:

Show 9 quoted lines

> On Friday 2007 March 16 10:05, Junio C Hamano wrote:
> > Andy Parkins <andyparkins@gmail.com> writes:
> > > Why does a bare repository have/need a HEAD?
> >
> > To mark which branch the repository owner considers the primary
> > branch of the repository.  I think "git pull $URL" (i.e. without
> 
> Interesting.  So it's got a much reduced importance and sounds like it 
> is mainly as guidance for cloners, etc.

It's as much a guidance for cloners, as it is for Git in non-bare repos to know which branch gets updated upon commit.

HEAD was never more than a pointer to one (the "current") branch.

Ciao, Dscho