Re: Anti-collide patch

[Michal *pht* Svoboda <svobodam@xxxxxxxxxx>]

On Mon, 11 Jan 1999, Piotr Kucharski wrote:

> On Mon, Jan 11, 1999 at 02:08:20AM +0100, Michal *pht* Svoboda wrote:
> > damn, yeah this is true. it gets messed on the path.. maybe then the nicks
> > should have unique id's or something. 
> 
> Sigh. Read http://akson.sgh.waw.pl/~chopin/uniqueID.txt
> 
> > > Apart from that, with such "feature" in the evening you would leave
> > > channel with Beeth!chopin@jantar, Vesa!ruokonen@umpahpah, Fusion!fusion@tyrell,
> > > syrk!syrk@millennium to see in the morning Beeth!ruokonen@umpahpah,
> > > Vesa!fusion@tyrell, Fusion!syrk@millennium and syrk!chopin@jantar or 
> > > something like that. Multiply it by 30000. Havoc, IMHO.
> > well only if you guys go and collide yourselves :)
> 
> Oh yeah? 
> B->A travel is 1 second
> A->B travel is 6 seconds
> 
> time     A knows                B knows              direction and message seen by B
>  0       Beeth@1,Vesa@2         pht@3                
> (here comes SJOIN with one way lag, as usually)
>  1       Beeth@1,Vesa@2         pht@3                <- NICK pht@3
>  2       Beeth@1,Vesa@2,pht@3   pht@3                <- :pht@3 NICK Beeth
>  3       pht@1,Vesa@2,Beeth@3   Beeth@3              <- :Beeth@3 NICK Vesa
>  4       pht@1,Beeth@2,Vesa@3   Vesa@3               <- :Vesa@3 NICK thp
>  5       pht@1,Beeth@2,thp@3    thp@3
>  6       pht@1,Beeth@2,thp@3    thp@3
>  7       pht@1,Beeth@2,thp@3    thp@3                -> NICK Beeth@1
>  8       pht@1,Beeth@2,thp@3    thp@3,Beeth@1        -> NICK Vesa@2
> (finally sendq is eaten)
>  9       pht@1,Beeth@2,thp@3    thp@3,Beeth@1,Vesa@2
> 
> Second 2 is pht<->Beeth  nickswap instead of KILL.
> Second 3 is Beeth<->Vesa nickswap instead of KILL.
>  
> Not only nick changed at A, but (nick) desynch introduced between A and B.
> "only if we guys collided ourselves", huh?

ok, this is true, i didnt notice it at first. of course every hacks known
from nick-colliding seem to remain. but at least Beeth and Vesa don't get
killed. and about the desync, A should send also :Beeth@1 NICK pht for the
first collide and same about Vesa and the second collide. maybe we should
then think of some mechanism that will make lag transparent to such stuff.

pht