[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: RFC: OP cheating



On Tue, 17 Mar 1998, forcer wrote:

> >But I think it'd be a great idea if on the third/fourth/fifth nick change
> >within a minute (or some such period) the client was given a notice and
> >set +r(estricted).

> +r is a bit drastic, don't you think so? I mean, someone might legimitetly
> change his nick a few times, and suddenly be +r :]

Not too drastic. (At least compared to the autokill suggested earlier.)
Well, I can imagine a normal oldnick->funnynick->oldnick or perhaps
oldnick->funnynick->newnick->oldnick. The timelimit could be raised if
this becomes a problem (which I don't believe happens). Now that I think
about it more thoroughly I think a limit of five nick changes should be
just fine. But the timelimit could be for example 2 minutes. I wonder how
many nick changes the current nick change penalty let's you do. However, I
think one can currently do far too rapid nick changes. 

Besides, a *normal* user who comes to irc to meet friends and chat
doesn't, as far as I am aware of, change nicks a dozen times a minute and
cry on +r.. :)

> Better would be either a temporary +r (like, MODE nick :+r nick and after 3
> minutes MODE nick :-r nick), since it will stop nickflooders, and won't
> hurt "accidently" nick"flooders" much ....

Hmm.. After the clones flood that period of time, they leave the channel
wait 3 minutes and go back? Worst nick floods go on for (a few) dozen
minutes where a few clones stopping for 3 minutes doesn't help - the
clones could just join/part flood that time. While join/part flood is
easier to stop and doesn't do as much damage to civilized(tm) irc clients,
it should at least be made sure they cannot use both.

By the way.. Should some more efficient channel cycling penalty be
implemented or would it already be too annoying to the users ?


> Besides, speaking of those "attack-types", lately there was alot of
> collide kills using lagged servers. The ircd2.9.* usage of "temporarily
> not available" to prevent that type of attack doesn't help here.
> In short, all the "attacker" does is connecting to server A (near the
> victim), and to server B (across the net and lagged), and when the victim
> changes nick, the client on server B changes to the same, and *boom* :)
> A single change would help here, namely the adding of timestamps on nicks,
> and then we could add them on channels etc. as well. Just an idea, though.

Err, you seem to be suggesting we'll just change to the efnet ircd while
at it ;-)

IMHO something (but what ?) should be done about the lag collisions you
mentioned.. What have the other non TS nets done about this ?


> 	-forcer

- Ville/viha@xxxxxx