gnu-arch-users
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Gnu-arch-users] Re: Re: [arch-users] advanced usage advice: the pri


From: Andrea Arcangeli
Subject: Re: [Gnu-arch-users] Re: Re: [arch-users] advanced usage advice: the prism technique (fwd)
Date: Sun, 28 Sep 2003 12:20:42 +0200
User-agent: Mutt/1.4.1i

On Sun, Sep 28, 2003 at 09:21:13AM +0100, Paul Hedderly wrote:
> On Sat, Sep 27, 2003 at 01:40:18PM +0200, Andrea Arcangeli wrote:
> > 
> > > about adding an argument to `tla get' so that it would hardlink instead
> > > of copying files from the revision library (if possible)? It escapes my
> > 
> > that would make it. Of course I'm playing risky, but I'm fine to deal
> > with that risk, the payoff makes it worthwhile, it's simply up to me
> > (not to the kernel since the kernel can't do that) to always do
> > copy-on-write on all the checked out trees.
> 
> Would it be worth tla making the checked out files read-only. That way

agreed, this will certainly make it safer and it sure work this way,
good idea.

> if you try to edit it you get a warning at least. You would need to do a
> "tla unlock ..." to edit it (which of course would rename and copy the
> file, maybe restore original permissions.)
> 
> Just trying to think of ways to make this safer... :O)

;)

tla unlock may even not be necessary (though it certainly wouldn't
hurt but I prefer to prioritize things and only add more worthwhile
things first ;). My rationale is  that everybody enabling the "total
hardlinking across cacherev and all the checked out trees" behaviour
will never ever run an unlock anyways after they get used to the full
speed (literally 1 second for the base-0 checkout instad of 1 minute,
you're not going to want to wait 1 minute anymore for your whole life,
that's why I consider this as a showstopper matter to checkout hundreds
of pure trees ;).

If you really want to do a not hardlinked checkout you can simply
disable the full hardlinked cacherev (or revision library, depends on
the implementation details), commit, delete the tree and check it out
again.
 
Andrea - If you prefer relying on open source software, check these links:
            rsync.kernel.org::pub/scm/linux/kernel/bkcvs/linux-2.[45]/
            http://www.cobite.com/cvsps/




reply via email to

[Prev in Thread] Current Thread [Next in Thread]