Re: LockDir

[Derek R. Price]

I didn't apply this:

KOIE Hidetaka wrote:

> In doc/ChangeLog,
>     >2001-06-08  Larry Jones  <address@hidden>
>     >
>     >        * cvs.texinfo (config): Mention using LockDir on in-memory
>     >        filesystem to speed up locking.
>
> In doc/cvs.texinfo,
>     >It can also be used to put the locks on a very fast
>     >in-memory file system to speed up locking and unlocking
>     >the repository.
>     >You need to create @var{directory}, but
>     >sc{cvs} will create subdirectories of @var{directory} as it
>     >needs them.  For information on @sc{cvs} locks, see
>     >@ref{Concurrency}.
>
> I comment:
>
>     Take care owner and permissions of @var{directory} and
>     @address@hidden/..} as same as CVSROOT.

This is already covered by (from cvs.texinfo):


     Put @sc{cvs} lock files in @var{directory} rather than
     directly in the repository.  This is useful if you want
     to let users read from the repository while giving them
     write access only to @var{directory}, not to the
     repository.

>     If using in-memory file system, it is volatile during reboot,
>     don't forget to create @var{directory} in a startup script like /etc/rc.

This seems better to leave to the filesystem docs.  Not only should any user
capable of setting up an in-memory filesystem be capable of solving this
problem on their own after at most a single reboot, but I can conceive of a
disk-backed in-memory filesystem that doesn't have this requirement.

Feel free to argue the point, of course, but I think I will wait until the
question comes up a few more times on this list to move the answer to the FAQ
and maybe later to the Cederqvist.

Derek

--
Derek Price                      CVS Solutions Architect ( http://CVSHome.org )
mailto:address@hidden         CollabNet ( http://collab.net )
--
... one of the main causes of the fall of the Roman Empire was that,
lacking zero, they had no way to indicate successful termination of their C
programs.

        - Robert Firth