Re: RFC: OP cheating

[Christophe Kalt <kalt@xxxxxxxxxxx>]

On Mar 16, Donald Quixote VI wrote:
| 1]	When an op quits, for any reason, lock out the nick for several minutes,
| perhaps 5 to 15 minutes.  5 minutes would probably cover 90%+ of op
| cheating attempts and go along way to making it an ineffective and
| unattractive means to takeover.  This being eliminated will go a long way
| to stopping clone floods as tsimple floods are easily twarted.

#defin'ing BETTER_NDELAY in config.h will lock out all
remote nicks that quit for 15 minutes.
it's the old style nick delay, quite impopular.

what you want is something in-between which is actually easy
to implement.  I wonder how unpopular it would be.

| 2]	When an op quits [perhaps within 5 minutes of being op?d], for any
| reason, send -o modes to that nick every several seconds for a period of a
| minute or so to cancel any +o modes that may be floating on lag.

this doesn't make much sense. think of it from the op target
point of view, rather than from the opper point of view.

| 3]	For nick changing clones, monitor nick changes and if a client changes
| nicks more than say 3 times in one minute, set that client to +r,
| restricted for the duration of that connection [ or kill it] -- they will
| prolly resort to join/part flooding and or ctcp although all but the
| newbiest know how to protect against ctcp abuses.

as you point out yourself, this won't take us far.

Comments anyone?