[off-topic] Server-side IRC channel logging? (not statistics)

[Joshua Isom]

On 8/18/2010 8:51 PM, Glen Barber wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> Hi,
>
> I recently set up an IRC server (irc/ircd-hybrid), which I don't see
> obvious settings for finely tuned channel logging.  What I would like to
> do is log individual channels without depending on a connected client.
>
> In all my searching I found software that either:
>
> 1.) depends on a 100% connected client, but provides concise logging of
> channel activity;
>
> 2.) logs statistics, rather than the useful information I am trying to
> obtain such as, who pastes the most links, who 'smiley's the most, etc.
>
> My interest is in the useful information in the channel, not statistics;
> ultimately, I want to have the channel conversations archived.
>
> I'd like to do this on the server itself.  For example, in the event I
> have to reboot my machine or the disk dies, or whatever bad event, I
> don't want to concern myself with missed data, corrupt logs, or a
> disconnected client, so I would like this to run unprivileged and
> without an interactive shell.
>
> If anyone has any suggestions, I'd be happy to hear them before I go
> reinventing the wheel.
>
> Thanks, best regards, and sorry for the off-topic post.
>
> - --
> Glen Barber

So you can set up the server but you can't install a client on the 
server machine?  If you put a client on the server, then if the server 
goes down, the client goes down anyway, but if the server goes up and 
you make a file in /etc/rc.d/ then the client also goes up.  You can 
have "near" continual monitoring, assuming you have a stable client and 
you're not concerned about those seconds when the server comes up.  If 
you want it done on the irc server you'll have to find a irc server that 
can handle it.  But there's plenty of irc bots out there that can 
probably do everything you want it too and if it's installed on the 
server hardware you'll have as good a reliability as the server itself, 
as long as the logging is good.  If the disk dies, data dies, that's 
just the way of life.  You could mitigate that, but it's always a 
possibility.