Re: Fwd: [CF:1317] client 0.95 dies with "Got error on read", other probs

[Frank McKenney <frank_mckenney@mindspring.com>]

> I've done a fair amout of playtesting of the current snapshot,
> and I believe that the current CVS snapshot is much more stable
> than our average release.

A compilation devoutly to be wished!   (;-)

--(assorted good news re bug fixes snipped)--

> > The frustrating thing is that 0.93 (486DX4-100 server/client, P5-100
X
> > server) was remarkably responsive over my 2-NICs-and-a-hunk-of-coax
> > "LAN", and I don't know where to start addressing the 0.95 
> > sluggishness.
> 
> You said you're using the new (better) graphics?  The better graphics
> are unfortunately more expensive in both network and CPU.

Well, I'm using -xpm on CF 0.95 as I was on CF 0.93, and the end result
is a 256-color rendition, but I have to admit I don't know if that
reflects the actual network transfer volume. You all _could_ be sending
32-bit color from the server to the client (process to process), and then
I force the poor client to squish that down to 256 colors and send it out
over the LAN to my X server.

> A while ago other crossfire hackers split crossfire into a client
> and a server.  The server controls the game but does no graphics.
> It instructs the client on what objects to display and animate.
> The client now does all the graphics.

Gotcha - the server sends the basic information, and the client then
renders it in whatever form seems appropriate for the client's platform. 
The only "X" in _my_ case, then, is the 0.95 client talking to my OS/2 X
server.

Thanks...

Frank



Frank McKenney, McKenney Associates
Richmond, Virginia / (804) 320-4887
E-mail: frank_mckenney@mindspring.com