Client-side a-priori moderation

[progval]

As far as I'm aware, there are currently three ways to deal with spammy messages:

1. block them at the server level before they are relayed to other servers/clients (eg. Inspircd does this)
2. the traditional way, which is to let them go through and have a bot quickly ban/kill the spammer
3. the upcoming edition/deletion specification.

1 requires server support and isn't very customizable by channel ops, 2 has obvious downsides, and 3 needs client support to delete spammy messages.

So I would like to propose a new option: a priori moderation by clients.

It would work this way:

1. when a client sends a message to a channel with such moderation enabled, messages are only sent to ops (like Charybdis' +mz modes)
2. ops can ACCEPT or REJECT these messages, using msgid. This would most likely be done by some kind of bot.
3. probably a fallback if ops do not accept in time

Potential issues:

1. increased delays
2. incentive to DoS the moderation bot to either delay messages further or fail open/fail close

Thoughts?

[nillkitty]

This would work really well in combination with a specification for "auditorium mode" which IRCX had (+x) and many server impl's have some concept of.

My preference would not to reserve the ACCEPT and REJECT commands for solely this purpose however as those are relatively broad command names that may have other existing or future uses. There should be a monolithic command for manipulating messages in various ways. There will be other ways in which moderations will need to manipulate messages in the future.