32 messages in net.sourceforge.lists.courier-maildrop[maildropl] Re: [courier-users] Re: O...
FromSent OnAttachments
Sam VarshavchikJan 13, 2003 3:46 pm 
D. J. BernsteinJan 13, 2003 6:11 pm 
Sam VarshavchikJan 13, 2003 9:11 pm 
Russell NelsonJan 13, 2003 9:46 pm 
Sam VarshavchikJan 13, 2003 10:19 pm 
Russell NelsonJan 13, 2003 11:11 pm 
Sam VarshavchikJan 13, 2003 11:35 pm 
mw-l...@csi.huJan 14, 2003 7:40 am 
Sam VarshavchikJan 14, 2003 3:22 pm 
mw-l...@csi.huJan 14, 2003 11:13 pm 
Sam VarshavchikJan 15, 2003 5:11 am 
Matthias AndreeJan 15, 2003 9:55 am 
Matthias AndreeJan 15, 2003 12:59 pm 
Matthias AndreeJan 15, 2003 1:36 pm 
Sam VarshavchikJan 15, 2003 3:11 pm 
Matthias AndreeJan 15, 2003 4:13 pm 
Sam VarshavchikJan 15, 2003 4:47 pm 
Johan LindhJan 15, 2003 10:16 pm 
Peter C. NortonJan 15, 2003 11:52 pm 
Bill MichellJan 16, 2003 1:30 am 
Johan LindhJan 16, 2003 2:00 am 
Bill MichellJan 16, 2003 2:28 am 
Matthias AndreeJan 16, 2003 2:28 am 
Matthias AndreeJan 16, 2003 2:45 am 
David LaightJan 16, 2003 3:14 am 
Sam VarshavchikJan 16, 2003 5:01 am 
Johan LindhJan 16, 2003 6:28 am 
Matthias AndreeJan 16, 2003 9:47 am 
mw-l...@csi.huJan 16, 2003 12:48 pm 
Sam VarshavchikJan 16, 2003 2:55 pm 
mw-l...@csi.huJan 17, 2003 12:30 pm 
Matthias AndreeJan 18, 2003 5:05 am 
Actions with this message:
Paste this link in email or IM:
Paste this link in email or IM:
Atom feed for this thread
Paste this URL into your reader:
Subject:[maildropl] Re: [courier-users] Re: OpenBSD 3.2 breaks Courier, Qmail.Actions...
From:Matthias Andree (matt@gmx.de)
Date:Jan 16, 2003 2:28:43 am
List:net.sourceforge.lists.courier-maildrop

On Thu, 16 Jan 2003, Johan Lindh wrote:

I actually had another brainstorm today. Add the file's inode, when moving the message from tmp to new:

tmp/timestamp.unique_pid.hostname -> new/timestamp.inode.unique_pid.hostname

Well, if the goal is to eliminate race conditions, this still won't do the
trick. Remember one of my early mails about this, Sam? Before we took it to the list.

The race condition causes a collision in new/* vs. cur/*. As long as the tmp/* name is unique (remember the maildir strategy backs off if there is a collision in tmp/*), there is no race.

I suggested making one of the always-present daemons a central point of message names.

Have courierd (for example) simply hand out unique file names on a TCP port. Simple to do, and completely, absolutely, eliminates all race conditions and conflicts. You no longer need PID's, inodes or whatever.

You'd essentially have to make all MUAs that are maildir() aware use the same daemon when for example moving mail from one Maildir to another. This adds networking code to things that don't need it, so it won't happen.

OTOH, this might be of more universal interest, and would best be placed in the kernel.