Wikipedia:Huggle/HAN

HAN stands for Huggle Antivandalism Network.

It is meant to let users of any automated tool, such as Huggle, share their activities, so they operate together in more efficient way. HAN was introduced in Huggle 3.

What it does

It shares basically everything you do in Huggle, when you revert and edit, you tell that to other users so that they know someone already attempted to revert it so they don't do the same, and prevent conflict. When you send a warning to user, other users can know it and increment the warning level even without having to reparse their talk page again (which isn't efficient because vandals can blank it). You share your badness scores, suspicious edits and similar information so that all users can work more efficiently.

How it works

HAN is currently using IRC protocol to communicate with other users, because it's very simple and effective. Current HAN lives on irc.tm-irc.org.

Why it is on tm-irc.org

Unlike freenode and similar ancient networks, tm-irc comes with features like auto-cloaking (new users have hidden IP even if they are not registered to network), and far higher limits (users can easily send a huge number of messages, reconnect frequently, use more sessions to connect and join nearly unlimited number of channels). Freenodes comes with a lot of restrictions that would make it very hard to enable such a thing.

Developers

Connecting

IP: irc.tm-irc.org
Port: 6667

To connect via web, go to https://chat.tm-irc.org/ enter a nickname, and connect to one of the channels for a specific Wikipedia, such as the English one which is https://chat.tm-irc.org/

Message Format

There is always a channel for each wiki, for example enwiki has #en.wikipedia.huggle, every message that is sent as plain text is considered a chat message and is displayed in chat window, commands are prefixed with double bold (char)2 symbol.

Commands are bellow:

• ROLLBACK <revid> - information that edit with id was rolled back by you
• GOOD <revid> - information that edit with id is a good edit (not a vandalism, shouldn't be reverted)
• SUSPICIOUS <revid> - the edit with id seems to be suspicious to you, but you didn't revert it
• WARN <level> <user> - the user was warned by you with warning of level that you used
• SCORED <revid> <score> - used by bots only - give the certain revid a given score