Re: Git hosting

["Elijah Newren" <newren gmail com>]

On Wed, Mar 5, 2008 at 8:46 AM, Olav Vitters <olav bkor dhs org> wrote:
>  > elsewhere with such a checkout, allowing users to essentially operate
>  > in a centralized mode.)  While this sounds cool, it really isn't all
>  > that useful because:
>
>  Meaning Git can do it, but it is painful, yet again?

No, actually, it's pretty easy:

  git clone --depth <depth> ...

I just don't see the point.

>  > Your assertion that shallow checkouts saves lots of diskspace is not
>  > well founded in my experience.  It was certainly true of bzr when bzr
>  > originally first implemented the feature, but these days systems are
>  > able to get a full copy of the history plus a working copy and STILL
>  > often take less space than an svn checkout.  Source code and its
>  > changes over time compresses really well with smart algorithms.
>
>  You're comparing to SVN?

Well, that is what GNOME currently officially uses...  :)

>  checkout + history is always going to be more than checkout only. E.g.
>  for sysadmin purposes I care about having a checkout, being able to
>  update that easily (without any possible merge errors), but I really do
>  not care about history / diff / etc. Complete waste of diskspace.
>
>  SVN is crap because it stores each file twice (for the diff).

Interesting.  I loved that part of svn compared to cvs because I
thought it was an awesome tradeoff; fast diffs are incredibly
useful...in the cases I'm most familiar with.

Always good to hear more usecases that are different.  Thanks for
pointing this one out.

>  Another question: I know git needs 'git gc' and that should be done
>  automatically. Are there other need to knows? Especially on the server
>  side? Is there stuff available to do a full mirror? Assume it should be

My server side knowledge is weak, sorry.  (My gut reaction was "why
would there need to be special software to mirror the git repos?" but
I really just don't know much of the sysadmin side of things.)

>  possible due to DVCS, but still.
>  E.g. Canonical goes down, and we just change the IP to the Red Hat
>  server. I know that due to DVCS people can continue to work.. but
>  still.. backups should be done anyway.. things should be up and running
>  in hours max (shorter is not possible with volunteer sysadmins).

I agree.

>  > I will note that a subset of the functionality from git checkout and
>  > git reset often get confused with git revert.  Another nasty unneeded
>  > UI wart (that is surprisingly easy to fix I might add...).
>
>  Still magic to me... maybe I'll understand more after tonight (trying
>  commands from behdad).
>
>  For Git, we'd need to hire someone to fix it up.

Or wait for a sufficiently motivated volunteer to stir things up.  :)