Follow-up Comment #3, bug #18664 (project freeciv):
Ah, well, that wouldn't be a problem :)
What I'm worried about is the rush of events at turn change that means there
might be a significant delay between the server starting a timer and sending a
packet and the client processing that packet and starting its timer. I suspect
that can cause discrepancies with the server and client's view of the timeout
today (even ignoring network transmission delays, which we can probably assume
to be small for Freeciv to be playable at all).
However, thinking about it, it might not be so bad. The natural time for the
server to send the client the action_timestamp-based information would be when
sending the packet updating the unit just after the user has moved it, when
there might not be much else going on (no flood of traffic from AIs). Unless
the client's not going to process it until it's finished animating a long goto
chain, or something. Hm.
Given these restrictions, I suspect it's not a good idea for the client to
use this information to prevent the user attempting to move a unit; it should
just display the icon (which could be a multi-stage countdown type of icon if
we're feeling keen).
Reply to this item at:
Message sent via/by Gna!
Freeciv-dev mailing list