We use an IRC channel that's on freenode. Feel free to join. The needed information is as follows:
Etiquette and fun
- It's impolite to store logs of this channel. Please disable logging before joining #noisebridge.
- It is not impolite, however, to add funny quotes to the quotes page.
Unfamiliar with IRC?
Here's some good general documentation about IRC and Freenode.
For those who are not familiar with getting on IRC, here's what Mitch did:
- For Windows: Download and install Pidgin (free IM chat software for Windows -- there's plenty of other free software to choose from for Windows and other OSs).
- For Macs - Colloquy (that also supports SILC) is Free Software Colloquy.
- For those who run Ubuntu and are not familiar with IRC (unlikely?): $ sudo aptitude install xchat
Then (using Pidgin on Windows):
- create an IRC account on Pidgin (good documentation for this step)
- Go to Accounts (I got to Accounts by right-clicking on the icon in the system tray)
- Click the Add button
- For the Protocal, choose: IRC
- For the Server: irc.freenode.net
- If you have a Screen name (a “nick”), type it in as your Username (If you don’t own one yet, make one up)
- Click on the Advanced tab, and fill in the Real name field
- Check the Use SSL box, if you like
- Click the Save button
- join channel: ##noisebridge:
- In the "freenode-connect" tab that appeared in Pidgin, type: '/j ##noisebridge' (without the quotes)
- So that no one else can use your nick, you can register your nickname at freenode.net
Hacking on IRC
Some members only want IRC to be used to discuss hacking. To satisfy this we have developed a system to ensure that all discussions are about hacking.
- Say, "I am wondering about (topic)"
- Then say, "Can we do some (topic) hacking?"
For example, "Wow, can you believe that (recent drama) happened in the space" is not correct. The correct approach would be, "We really need to do some drama hacking. Can you believe that (recent drama) happened in the space?"
Congratulations, you have just been excellent to other members by only talking about hacking in the IRC channel.
Occasionally, people and programs may generate traffic in the #noisebridge IRC channel with the intent of provoking negative emotional responses. If you currently have operator status in the IRC channel when this happens, use the following guidelines to handle the situation:
- Do not op users if they have not authenticated with NickServ. Check if a user is authenticated with NickServ by sending NickServ the command info <user> in a private message.
- Consider setting +r (only allow users registered with NickServ) and -t (only ops can change the topic) on the channel. The overall channel mode should be +npr in this case. Under normal circumstances the channel mode should be +npt.
- Consider using +q (channel-level ignore) instead of +b to silence the provoking users. This allows for users who were mistakenly caught with a hostmask that was too broad to still join the channel and find an op to help troubleshoot.
- Remember that the motivation of the provoking users is to generate emotional responses. If they do not see any, they will move on to a more responsive channel. It is important to avoid talking about them.
There is an inherent conflict between allowing free speech in #noisebridge and restricting the ability of trolls to disrupt the exchange of ideas. Any restrictions on speech must be temporary. Remember to undo any changes you make within about 24 hours.
Some tips on creating hostmasks:
- Around trollcon 2, trolls may start joining from multiple domain names. Make sure to reverse lookup the names to see if they're in the same network block. If they are, set a mask for the whole block. For example, morb (an actual troll) has control of several machines in 220.127.116.11/29. Two of the machines are named failure.mysqrl.com and variance.misentropic.com. Looking up these names shows that they're in the same network block:
$ dig +short variance.misentropic.com 18.104.22.168 $ dig +short failure.mysqrl.com 22.214.171.124
Setting a mask of *!*@208.122.47.* will catch all computers in the surrounding /24 without having to enter masks for each one individually. For more precise information on what network block the addresses are in, try using a referral-following whois tools like whois.domaintools.com.
- Don't bother with username masks (*!email@example.com) unless you know that a troll is on a shared vhost they don't have root on (such as chef.nerp.net). It is common for trolls to be able to spoof ident replies.
- Note that afraid.org provides a large pool of domains which can have reverse DNS entries added by arbitrary members of the public. Subdomains under a domain hosted by afraid.org may be entirely legitimate, although you should report offensive subdomains to dnsadmin at afraid dot org for TOS/AUP violations.
Recommended channel mode: +Cn
People are chatting about their business as usual, occasional frustrated ranting by someone.
Recommended channel mode: +Cnpt
Troll present and attempting to provoke an emotional response.
Procedure: Set +q on the troll.
Recommended channel mode: +Cnptj 1:60 and possibly +q $~a to quiet unregistered users, or even +r to prevent unregistered users from joining at all.
Multiple trolls incoming and spamming the channel.
Procedure: Set +j 1:60 on the channel to throttle joins and buy a little time, then start setting +q on the trolls. If the trolls are coming through a web gateway or an IRC bounce host like xzibition, then use a forwarding ban similar to this:
If the spam makes it hard to see or cut and paste names to create hostmasks, then quiet all nonregistered users temporarily:
Recomended channel mode: You're boned. If possible, use a kickban everyone script to buy time.
Channel takeover, troll(s) have ops, spambot(s) spewing text into the channel.
Procedure: Get a hold of radii, Syzygy-, or Valkyrie` to override ops and reclaim the channel. "/msg chanserv access #noisebridge list" will give you the current list of registered channel operators.
It's not guaranteed, but Freenode staff may show up and help deal with it, as they claim to have a mechanism for detecting these kinds of problems.