I have implemented auto-enable now.

Here's a console transcript showing the new functionality.

(#:~/mirrorbrain)- mb/mb.py disable ftp4
(root@mbwork)(1580/pts/3)(05:27pm:09/06/10)-
(#:~/mirrorbrain)- mb/mb.py show ftp4 | grep statusBaseur
statusBaseurl  : False
(root@mbwork)(1581/pts/3)(05:27pm:09/06/10)-
(#:~/mirrorbrain)- mb/mb.py scan -e ftp4                 
Mon Sep  6 17:27:43 2010 ftp4.gwdg.de: starting
Mon Sep  6 17:27:43 2010 ftp4.gwdg.de: total files before scan: 1109
Mon Sep  6 17:27:44 2010 ftp4.gwdg.de: scanned 1109 files (1372/s) in 0s
Mon Sep  6 17:27:44 2010 ftp4.gwdg.de: files to be purged: 0
Mon Sep  6 17:27:44 2010 ftp4.gwdg.de: total files after scan: 1109
Mon Sep  6 17:27:44 2010 ftp4.gwdg.de: purged old files in 0s.
Mon Sep  6 17:27:44 2010 ftp4.gwdg.de: now enabled.
Mon Sep  6 17:27:44 2010 ftp4.gwdg.de: done.
Mon Sep  6 17:27:44 2010 ftp4.gwdg.de: testing status of base URL...
Mon Sep  6 17:27:44 2010 ftp4.gwdg.de: OK. Mirror is online now.
Completed in 1 seconds
(root@mbwork)(1582/pts/3)(05:27pm:09/06/10)-
(#:~/mirrorbrain)- mb/mb.py show ftp4 | grep statusBaseur
statusBaseurl  : True


I think that this is the best path to go -- following the principle of least surprise.

Committed in r8116.
http://svn.mirrorbrain.org/viewvc/mirrorbrain?view=revision&revision=8116

Marking as resolved.

After enabling a new mirror (mb scan -e), it is not really enabled, because it 
first needs to be marked "online" by the (separately running) mirrorprobe. This is 
something that new users frequently run into.

The mb scan command could mark the mirror online right away. The only reason not 
to do so is to prevent harm from typos, and from the mirror going offline. 

If the mirrorprobe is already configured to run every minute, it would disable a 
broken URL soon anyway. 

Users may miss to set up the mirrorprobe in general; but forcing them there by the 
above situation is not really nice.