Re: Could svn check permissions?

e.huelsmann@gmx.net wrote:

> Patch bdb? ask them nicely to do it differently?

I still maintain that there is nothing wrong with BerkeleyDB. As long as you
set up the directory, groups, and file permissions correctly, it Just Works(tm).

How about this for a solution:

1) Add a new option to 'svnadmin create'

   -r [--revision] arg : specify revision number ARG (or X:Y range)
   --group=ARG : create repository as group ARG

which will set the appropriate rights to start with, rather than requiring the
site admin to do it manually;

2a) Alter svn to always set umask S_IRUSR, S_IWUSR, S_IRGRP, and S_IWGR prior to
opening the database (will this work?);

or

2b) Include a svn wrapper which sets umask 002.

>
> Anyway, if it were a non-issue, then why does it arrise so often on the
> lists?

Mostly it's because the file:// access is so much faster than http:// and
because so many people are used to using :local: mode with CVS. I view it as a
combination of a documentation failure, historical baggage, and the odd nature
of shared file access via a mediated server process library.

John

-- 
John Peacock
Director of Information Research and Technology
Rowman & Littlefield Publishing Group
4501 Forbes Boulevard
Suite H
Lanham, MD  20706
301-459-3366 x.5010
fax 301-429-5748
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org