Custom Query (1145 matches)

Reported by geofft, 13 years ago.

This is a followup to ticket #229, which was resolved for cluster machines with cluster-cups-config, hardcoding a list of printers.

However, less-than-cluster machines still at the moment browse for printers against cluster-printers.mit.edu, and as established in that ticket, this means they also implicitly browse the local network. As long as printing exists in some form, there will exist misconfigured local machines re-advertising the Athena print queue(s) through themselves.

We should find a way to distinguish real Athena print queues from similarly-named print queues on other servers, without necessarily breaking the option to scan for printers on the local network if the admin wants to (as comment #12 alludes to). I don't have a strong preference as to whether it should be on or off by default — if we can fix the upstream CUPS bug that is unintentionally forcing it on, great.

One approach would be to prevent any other queue with a name matching the Athena print queue(s) from showing up, but let all other queues through. (This actually vaguely reminds me of nss_nonlocal, come to think of it.)

Reported by jweiss, 13 years ago.

installing the debathena-discuss-server package ended up with discuss in the password file using a random UID, and not 32000 (the UID discuss has in hesiod/moira/etc, and the one I expected it to use).

As an aside, the UID assigned happened to conflict with something else in hesiod, but I understand that's a hard problem, and am willing to ignore it.

Reported by geofft, 13 years ago.

The current public source repository, /afs/dev.mit.edu/source/svn-repos (served via  svn://debathena.mit.edu/athena), is updated periodically by a cronjob that has read access to the real repository at svn.mit.edu:/svn/athena. This means that people reading commit zephyrs won't be able to look at the commit immediately, nor will Trac show the commit in the browser immediately. The same is true of the public source checkout in the "source" locker.

It would be nice to have this updated more quickly. There are two possible solutions I see:

1. Update the mirror and the checkout from a post-commit hook. Greg Hudson says that the reason this currently doesn't happen is that post-commit hooks don't get environment variables, and therefore don't have tickets to use to "aklog dev". When we move to the athena cell, this may become much easier. You could also imagine some process involving calling a CGI to go do the fetch, or something.
2. Have a svnserve pointed at the real /svn/athena. Is this what anonsvn.mit.edu is for?