On Tuesday 17 July 2007, Per Inge Mathisen wrote:
On 7/17/07, Martin Koller [EMAIL PROTECTED] wrote:
What about the follwing patch ?
Just a quick glance at src/multibot.c tells me you missed cases. That
will break multiplayer games badly, so it is important you catch them
all.
Do I see
On 7/18/07, Martin Koller [EMAIL PROTECTED] wrote:
Do I see this correct, that
NetAdd(m,4,psDroid-x);
NetAdd(m,6,psDroid-y);
adds the value psDroid-x/y at byte position 4/6 ?
So that means x/y must be 2 bytes to make it work ?
Yes.
If so ... what a code .. :-(
Indeed. It is on
On Wednesday 18 July 2007, Martin Koller wrote:
It should better use a size counter and sizeof() instead.
What do you think about the following approach:
A new NETadd macro simply counts on its own what is given to its NETMSG.
So e.g. sendHappyVtol() would change from:
NETMSG m;
On 7/17/07, Martin Koller [EMAIL PROTECTED] wrote:
What about the follwing patch ?
Just a quick glance at src/multibot.c tells me you missed cases. That
will break multiplayer games badly, so it is important you catch them
all.
See eg sendDroidDisEmbark(). I would grep for psDroid-x, psStruct-x
On Monday 16 July 2007, Per Inge Mathisen wrote:
On 7/16/07, Martin Koller [EMAIL PROTECTED] wrote:
I ask myself: what is the x,y,z position supposed to have as datatype ?
I find UWORD but also UDWORD.
..
I think this should be made consistent to be able to correctly fix the
compiler
On Monday 16 July 2007, Per Inge Mathisen wrote:
Nice, this removes a lot of ugly casts. Unfortunately, that patch
would break the netcode badly. In order to convert basedef's xyz from
UWORD to UDWORD, we will have to check or change the size of all xyz
coordinates going over the network, too