Re: GR proposal: code of conduct
-----BEGIN PGP SIGNED MESSAGE-----
- From the IRC perspective, we welcome the CoC -- it's nice to have
a document to point people to which clearly articulates the sort of
behaviour we want and the sort of environment we would like to
maintain. A CoC is a great aspirational device and having the project
support it sends a clear message.
Let's not start making this into a massive bureaucratic nightmare
for all of us though. We echo the concerns of Alexander Wirt on this.
Stefano Zacchiroli wrote:
> For IRC it's a bit more difficult, because we do not long our IRC
> channels by default (or at least I'm not aware we do), with the
> exception of meetings run with the help of meetbot. That means that it
> would be rather difficult for the moderators to point out to the
> evidence on the basis of which they've banned someone. I can't help
> wondering if the solution to this shouldn't just be radical,
> i.e. publicly log our IRC channels. A less invasive solution is to just
> ask moderators to publish log excerpts that they think justify the ban.
IRC bans are already public in that the mode change is public to anyone
on the channel at the time and anyone on the server can inspect the ban
list for a public channel. I assume that the above is written with the
developer channels in mind, but let me state that as an op in the project's
public facing channels (#debian, #debian-offtopic) I would strongly object
to having an increased workload put on the ops of those channels. The ops
there are already stretched  -- having to deal with crap in #debian
isn't the most fun way of spending your time in the first place and adding
a layer of bureaucracy to the process is just going to suck the life out of
Looking through logs for the second half of last year, I find that
there were almost 400 bans placed in #debian (oftc + freenode) and that
all but a handful of them were made by 3 hard-working volunteers.
Who is getting banned? In rough order:
* spammers who want to crapflood the channel with race hate
* other link spammers/scammers/jibberish bots
* people who want support for windows, redhat, ubuntu, mint ... and
refuse to actually ask in the right place where there is a dedicated
channel with people to answer their questions
* people with broken irc clients repeatedly joining then parting
* people who want to rant about something offtopic, monopolise the channel
by doing so and thus prevent anyone else from getting the support they
are seeking. 
How long are they banned? Usually a few minutes is enough to be able to talk
to someone in private, to get them to cool down and be constructive again.
If someone wants to set up some sort of paperwork scheme to track that
sort of information, they are are welcome to do so -- just don't expect
anyone to use it. They are welcome to idle in the channel and do the
paperwork themselves if they feel it is worthwhile. 
This is not to say that we want to do all this in secret and have no
transparency. We frequently discuss bans in our ops channel with other ops,
with network staff/opers and with people affected by them. More commonly,
the op in question discusses bans directly with the person in question in
a private message though as that's much more likely to de-escalate the
situation. All ops are able to remove bans they disagree with.
Time and people are routinely identified as the most important things Debian
lacks; let's make sure our effort is spent productively.
(with input from other #debian ops)
 before anyone suggests this, let me pre-emptively say that parachuting
new people in as extra ops doesn't work. New faces don't have the trust
of the regulars so they are ineffective ops. This has been tried before
and I don't believe there is one single active op left in #debian who was
 it won't surprise anyone that ranting about gnome 3 or init systems
is common enough... But #debian is a *support* channel; people there can
only answer questions and help people report bugs where necessary.
Anything beyond that becomes destructive to the channel and demoralising
to the people volunteering to help in there.
 yes, this is also an echo of other conversations we have had in the
project recently -- if someone thinks something is important, they should
step and do the work not require everyone else to bend to their wishes.
Stuart Prescott http://www.nanonanonano.net/ firstname.lastname@example.org
Debian Developer http://www.debian.org/ email@example.com
GPG fingerprint BE65 FD1E F4EA 08F3 23D4 3C6D 9FE8 B8CD 71C5 D1A8
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.12 (GNU/Linux)
-----END PGP SIGNATURE-----