#8 exempt string!

[Anonymous]

I'll like to ask if theres a command for nospam1.7.2,
which is to exempt some strings, lets say i had set a
particular string to *#*, it acts to anybody who does
say 'join #mychannel'.

But now the situation is when people with script joins my
channel, lets name my channel as #jingyou, they'll
say 'Thanks for voicing me in #jingyou', and so my bot
will start to kill that user.

I'll like to know how do i place a exempt string to
#jingyou.
Thanks for your time in helping out with your free time=)

[Philip Richardson]

Logged In: YES
user_id=113655

This is a definate somthing I will look to adding.

As explained in mail's, the egghelp.org forum and other
feature and bug reports. I have a lack of computing
equipment, which I hope to remidy within a month.

If I impliment this feature, it will be based on a user setting,
that would affect backwards compatability in no way what
so ever.

Each spam type (invite, message, channel, notice) would
have it's own new setting, with a possible 3 values
(depending ont he spam type).

0 = No exemption
1 = Exempt channel names, equal to the channel the
message was destines for
2 = Exempt channel names, for any channel the bot is
monitoring

In more detail

This feature is only effective, if you have some form of spam
string to puish users, saying channel names.

0 would mean, that no matter what channel, or spam type
used, the script will not exempt any messages. This would
be the default, as it is the current behaviour.

1 would effect the punishment, only if the channel name
said, is not the channel it was destined for. In items like
private messages and private notices, this would be the
same as 0. In the case of a invite, it would be the same as
2.

2 would effect punishment, only if the channel name said, is
not being monitored by eggdrop. IE, if not listed in the
output of .status.

In all cases, execpt setting 0, this could be a rather simple
way around spam detection.

IE, sending a message to channel #help saying "#help -
Why not join #sexxxx" would defy protection.

The same applies for private messages. Most spam bot will
be sending message to the no-spam bot, after detecting it
in a channel. As such, prpending the message with the
channel name, would defy the protection.

There is no simple way around this, without the inclusion of
some form of AI (SEE: AI below). Which would bloat No!
Spam up no end.

Please reply with any coments or issues you have
regarding the way I wish to impliment this.

AI is the automated reaction, to a string of text. The
outcome of this reply, is limitless, but is heavily based on
the input string. Using pattern detection, a application
must scan a string, and detect the pattern and context a
string was used in. In many cases, a string would be
scanned multiple times.

In the case of no-spam, this task would more that tripple
the load of the script, creating a DOS prone script.