[1334] in Coldmud discussion meeting

root meeting help first first in chain previous in chain previous next next in chain last in chain last

Re: [COLD] Exit announcements (anomaly?)

daemon@ATHENA.MIT.EDU (Thu Aug 21 19:55:00 1997 )

Date: Thu, 21 Aug 1997 19:51:48 -0400
From: Chris Williams <psion@geekspace.com>
To: Brandon Gillespie <brandon@roguetrader.com>
CC: "Luther, Clay" <clay@selsius.com>,
        "'coldstuff@cold.org'" <coldstuff@cold.org>

> > I've noticed the same thing, and agree. Is there a reason source and dest
> > announces can't be called bevore move_to (in $exit.invoke)? (they'd both have
> > to be moved, else the actor would see their messages twice, or none at all).
> Yup, there is.. what happens if .move_to() doesn't allow you to move?
> Then everybody has been notified that you went/arrived, but you didn't...
Yeah. If I recall, MOO's approach to this problem was to approve the mover for
moving, then do stuff, then move them, but that was ugly. Or you could do the
approval checks twice, once just to check, and once as part of the move.
Maybe this should be handled by whatever's handling the CML? Like, as part of
the CML-frob, you give it meta-information about which messages should be
dispatched in what sequence, and to only do such-and-such messages on
condition of success of x verb-call (which the cml parser makes??). This seems
a bit heavy-handed.
Maybe this is just a case where two different messages are needed, instead of
piggybacking the actor's message on the destinations announce.

Version: 3.12
GCS/CM/IT/M/S d-- s+: a15 C++++$ UL+ P L-- E? W++(+++) N+ o? K? w++++@
O--- M-- V- PS++ PE- Y+ PGP- t+ 5++ X+ R+ tv+ b++ DI+++ D+
G++ e* h! r y-