[Freeciv-Dev] Re: packet batches? (was: [Patch] Making city report list faster)

[Christian Knoke <chrisk@xxxxxxxx>]

On Mon, Jun 10, 2002 at 10:15:36PM +0200, Reinier Post wrote:
> On Mon, Jun 10, 2002 at 08:51:32PM +0200, Christian Knoke wrote:
> 
> > > So adding an AI player decreases the amount of interleaving of client
> > > requests.  This means if A, B and C are playing, and A attacks B's city
> > > with 5 units at once, A's has a much better chance of success when C is
> > > an AI player!
> > 
> > But only with a lagless network and client AIs for A and B.
> > 
> > I understand that you can send a bunch of howitzers to a city and
> > attack it a dozen times. In this case the city is conquered even before
> > B is aware of being attacked - due to net lag, no matter of client ai
> > or not. How am I wrong?
> 
> A needs time to move those howitzers.  Let's say that B will notice the
> attack after 1 or 2 howitzers and respond before the 2rd or 4th hit,
> assuming normal network latency and reaction speed.  If A releases 12
> howitzers at the same time (that is, a time B doesn't have any chance
> to respond.  With the existing client, A can only do this across turn
> changes (because the server really needs some time there).  With a
> hacked client, it can do it anytime.

I did understand that. But my point is, this cannot be avoided by a
change in the server processing of the client requests. What is the
normal ping time to a server over the ocean? 100 ms? Time enough
for a server to process a hundred moves.

> So take the rating system with a grain of salt :)

Robot fights :-))

> Reinier

Christian

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