[Freeciv-Dev] Re: (PR#2223) cma: server knowledge of island != client knowledge

["ChrisK@xxxxxxxx via RT" <rt@xxxxxxxxxxxxxx>]

On Thu, Jan 09, 2003 at 09:15:06AM -0800, Mike Kaufman via RT wrote:
> On Thu, Jan 09, 2003 at 06:48:18AM -0800, Raimar Falke via RT wrote:
> > 
> > The question how does the server tells this the client. We have (in
> > increasing amount of information send/leaked):
> >  - server informs the clients that the cities of the traderoute are on
> >  the same/different island.
> >  - server informs the clients that the tile where the two cities
> >  exists are on the same island
> >  - server sends information about missing tiles to the client so that
> >  the client can calculate this (cities on same island) by itself.
> 
> a combination of the last two I believe. Using an idea that Davide had, any
> city getting a trade route from another city should get to know the
> location and name of the other city. I suggest that if B creates a route
> with A, then the server reveals (fogged) B's tile (with a dumb city).
> 
> In this case, lets say B appeared off in the black somewhere, A's owner
> obviously thinks that A and B are on different continents. The solutions is
> to B's owner's knowledge. somewhere in handle_unit_establish_trade():

Why do you do this so complicated (and errorprone)? I really don't
understand what is the benefit from this. Just let the server decide.
You're talking of "generalized effects", what is this?

Christian

-- 
Christian Knoke     * * *      http://www.enter.de/~c.knoke/
* * * * * * * * *  Ceterum censeo Microsoft esse dividendum.