ignore away-msg from kicked users

cyberal · Post by **cyberal** » 2003-11-17 04:20

Yes, how 'bout that?
An annoying little thing.. users manage to send away pm to you just before they get disconnected.. I would like to be able to ignore those msgs

Should be fairly easy to implement (not that I know C++), but if I were to write it in Java it wouldn't be a problem I think

Todi · Post by **Todi** » 2003-11-17 04:35

I must say that I have never experienced this, ever, despite having kicked my fair share of users... I guess it's an advantage with having a slightly lagged hub

Sedulus · Post by **Sedulus** » 2003-11-17 08:18

this has never happened to me either, even though I did some mass kicking recently (users with pre 0303 dc++'es).

I think you should blame your hubsoft.
the kick-pm you send, plus the kick command easily fit into one packet (except perhaps when kicking more than one person at the same time)

if the hubsoft processed this batch (incoming buffer) in one go, this would not occur.

fixing this in DC++ is kind of problematic, as the client does not know if the user is actually getting kicked, or, if the away message is really an away message.
consider the scenario when you are kicking from the filesearch, and accidentally try to kick a SuperOP (which fails). if dc++ were to remember that the "user is kicked" and ignore all messages, you wouldn't see the shouting of the SuperOP, and get your ass whooped for not replying with a nice excuse pm :D

the easy solution is to use hub-dependent kick commands, that have the kick-pm message and the kick-command in one command (+kick <user> <message>)

cyberal · Post by **cyberal** » 2003-11-17 08:26

yeah okay, ofcourse you are correct... the under lying reason is a laggy hub.. and I do kick more than one user at a time.. that may make it worse.. Building a routine to handle the ignore seems easy to me, maybe not efficient but easy..

simply add the kicked nick to a list when you activate the kick command.. and have DC++ check against that list when recieving a pm (if the list contains any nicks).. the nick can be removed after 5 seconds or something.. so your SuperOP theory would not happen!