2011/1/27 Ansgar Burchardt <ans...@debian.org>: > > I just took a look at the original bug report and fixed some of the > issues.
Great, thanks! I agree that points 1-4 and 7 can be considered fixed. That leaves 5, 6, 8, 9 and 10. > 5,6) Still needs to be changed. Maybe move to it's own section about > access control. I think this section should be moved to at.allow(5). I will submit a patch against git separately. I don't think there's any conflict between specifying that an empty /etc/at.deny is the default, while shipping a non-empty one on Debian, but arguably it's simpler just to remove the bit about the default, as any user will need to check whether /etc/at.{deny,allow} exist on their system in any case. The first part of point 6 is already addressed by the existing documentation, but it will be clearer when moved to at.allow(5). The documentation clearly states that output is mailed, optionally using the command MAIL_CMD, which answers the second part of point 6, about syslog: at cannot normally use syslog, but you could fake up a MAIL_CMD which logged the output instead of mailing it. I don't think this needs spelling out in the man page. Point 10 will be addressed by moving the full documentation on at.allow to at.allow(5). > 8,9) Still needs to be changed. timespec is not user-friendly either :/ Point 8 can be addressed by simply deleting the word "exact" from at(1) :) I'll include this in my patch. Point 9 I will address in a separate patch. -- http://rrt.sc3d.org -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org