[dcc2] SIDs and connections to self.

[myndzi]

The other option is simply to ensure that all the required info
remains in the message. That way you don't have to "remember"
anything, merely react to incoming data. That may be the 'way to go'.
Sorry for the mass posting ;P

On Sun, 27 Jun 2004 00:13:21 -0700, myndzi <myndzi at gmail.com> wrote:
> 
> I don't think that's worth it. Better to just keep two separate sets
> of data - one for "incoming" requests and one for "outgoing" ones.
> 
> I drew up a bunch of other stuff earlier, I'll post more about it later.
> 
> 
> On Sat, 26 Jun 2004 17:27:41 -0400, Dave Johnson
> <dave-dcc2-org at iroffer.org> wrote:
> >
> >
> > myndzi writes:
> > > I ran into this annoying little ambiguity while working on
> > > implementing DCC2 stuff.
> > >
> > > It makes sense to keep a single consolidated list of nick,sid pairs so
> > > that you do not end up with two different connections (one initiated
> > > by you, and one by them) with the same SID. However when you do this,
> > > sending DCC2 requests to yourself becomes troublesome -- obviously the
> > > nick and SID will be the same.
> > >
> > > Since you can't really tell if a DCC2 Accept CTCP is for a transaction
> > > *you* initiated or one *they* initiated, there being two possibilities
> > > here (direct reply to a DCC2 Publish, or a reply to a DCC Accept), it
> > > seems that some thought should be put into this. This is another good
> > > reason to change the name of each CTCP or add some sort of state
> > > counter.
> > >
> >
> >
> > Ya, another problem.  Similar to the problem I mentioned in a post on
> > Apr 25th where I gave an example of a SID race condition that would
> > have the same effect.  The proposal I gave there was splitting the SID
> > into two fields, MYSID and YOURSID.  That would work for this too. See
> > that post for more information.
> >
> > --
> > Dave
> >
> >
> >
> > _______________________________________________
> > dcc2 mailing list
> > dcc2 at dcc2.org
> > http://six.pairlist.net/mailman/listinfo/dcc2
> >
>