Peter Mardahl wrote: > > > The only remaining serious bug I know of is "that map bug": the one > > > where you come out of an exit and end up in the sea. > > > > I'll re-look at the enter_exit code - a rewrite may be in order in any case. > > > > But even that bug seems pretty rare. > > It's quite annoying though. It's the one thing we "really should fix". > I don't understand that area of code very well, unfortunately. The harder part is knowing if this is really fixed. I recall that this isn't a really reproducible bug, so play could go on a lot without really knowing if it is fixed (unless there is a real obvious error in the code). > So 5M total for a complete server+maps+client distribution? And about 40M > of disk space uncompressed? You... realize, of course, how lightweight > those demands are compared to a great many games? Or to Netscape? Its lightweight if you have a home system or work system. I'm curious if you are at a university what type of space they may give you now days. > > I'll volunteer to put together RPMs of maps + clients + server: > binary versions. I will need a client which will allow easy startup > of the local server + metaserver support. I do not think making > a binary version of the server would be difficult. I do not > see why you think everyone would have to compile his or her own > server. What is the difficulty with shipping server binaries? No real difficulty. There are probably more dependancies than on the client. > Certainly. What I was proposing was a greatly expanded application of > random maps, such as a whole city with many buildings, all leading to > deep random maps. I can also design random maps such that they lead to > more random maps. I'm not sure how deep I would want the random maps to get - and the problem remains that if that is a common set of maps that lead to a place a whole bunch of people want to go, the maps that lead there will tend not to get reset. I do like the idea of unused houses leading to random maps. > Better than reasonable, it sounds great. Let's: > 1) finish up the Gods I think Darth Bob said he would work on this. Note that the granting of prayers and whatnot in god.c need to be re-examined for PR code. > 2) finish up PR things Anything major here, or is it mostly twiddling of values? > 3) fiddle with the clients a bit to support the metaserver I'll do that. > 4) and local-server-startup A simple solution for this would just be a shell script that runs the server, waits 20 seconds or something for it to finish loading, and then runs the client. > 5) find and crush bugs > 6) tune balance and go to 0.99, Those two just have to get fixed as they are identified.