Apparently a previous message I sent to the list didn't make it...resending. (Trying to keep line lengths reasonable, but this is on a web-based email client so God knows what will happen, and I even have my doubts about Him.) Thanks for you earlier reply John. I had fetched the source from http://www.netrek.org/repos/netrek-server. I've made my repository publicly available at: http://brent.veroquest.com/bmiller/netrek-repos/ I've rolled my porting changes into a patch, but I've only compiled and run it on athlon x86_64 linux 2.6.11 (RedHat FC4). It's not tested on any other platforms. I'm also new to running netrek servers, so I've only minimally tested the feature sets--I haven't even gotten t-mode on. I've only tested the configuration that the vanilla server unpacks with (plus my changes). The default robot works, but apparently not even 8 GB of addressable memory is enough to allow the default robot to beat a human driving a scout. Perhaps on a 8-core machine.... :-) I spent 5 minutes trying to set up a newbie server but wasn't committed enough to get it working; if finishing 64-bit porting really does have value I would be happy to spend more time on robots. It might save me a few minutes if someone can send me a sysdef file (or files) that exercise(s) some other robot(s). Regarding helping to run a training server: I would be happy to contribute code changes, but I don't have access to a server with a fast connection, so I don't think I'm equiped to run the server. For my experimental changes, I had no intention of running another public server. I'm just looking to get some people to try it so I can get some feedback. The motivation behind my changes were: (1) to allow new players to learn the ropes without being bait for "free beer" hunters, and (2) have minimal impact on game play. I think my current changes can be improved on, but I figure that I have to start somewhere and get feedback, so this is my first pass. My initial feeling is that there is potential behind the idea of "no kill credit for taking out a twink," but it remains to be seen if there is way to make it work. Thanks, Brent On Sat, Nov 10, 2007 at 09:11:01PM -0600, John R. Dennison wrote: > Wrapped to 78 columns. > > On Sat, Nov 10, 2007 at 08:34:40PM -0500, bdmiller.netrek at mailnull.com wrote: > > Hello galaxy, > > > > I am running a vanilla server with some experimental changes and I'm > > trying to figure out the best way to test the changes and see if they > > have any merit. My server is running behind a DSL connection, so I'm not > > sure if the best way to test it is to advertise my server, or try to get > > someone else to run my code for a trial game or two on another server. > > If anyone has any suggestions, I'd love to hear from you. > > Please do not list servers that are not on decent connections; > we have enough problems with player base splits as it is > without additional servers being listed, especially those > that are not on fast pipes :) > > > The changes I've made are: > > (1) Some porting to build on 64-bit Linux (AMD, Red Hat Fedora Core 4) and > > clean up warnings. > > This is good. Altho pickled was running on a 64bit platform > for quite some time, and base-practice is currently on 64bit > now. > > The largest stumbling blocks with 64bit that Karth has not > recently addressed that I am aware of are: > > 1) Robot code - bots will not enter when compiled as native > 64bit code; 32-bit needs to be used for the bots. > > 2) cambot does not work on 64bit for INL games. > > Chances are that if you did not do a darcs get from a public > source repo you do not have his 64bit changes and your codebase > is going to be pretty out of date, altho to be fair I have not > checked to see what's in the current tarball. You can take a > look at the REPOSITORIES file in the source kit for active > source repos. > > > (2) Try to reduce the rewards for killing twinks, so newbies can live and > > learn in relative peace without hurting their team as much. > > * Killing people who have no kills doesn't grant ability to carry. > > * Longest-surviving player on each team can always carry. > > * Killing an opponent who can carry grants ability to carry. > > * Partial credit toward carrying still granted for bombing, etc. > > * Ships enter the game with < 100% fuel > > As these changes alter game play significantly, I am > not fond of any of these changes on a true bronco server; > they may have some value on a training server, which we > sorely lack at the moment. Would you be interested perhaps > in working on such a beast? > > > The changes related to #1 are straightforward and if anyone would like to > > review them I can push a patch to any repository. > > > > The changes related to #2 are conditionally compiled based on a flag in > > config.h, and then enabled/disabled in the sysdef file. I don't plan to > > push these changes unless/until I can get some feedback and determine if > > they have any merit. > > If you have webspace available the best thing to do would be to > make your source repo publically available and the developers > can pull your changes down and review them. If you choose to > do this just send your repo's to the list. > > > > > John ---------- This message was sent from a MailNull anti-spam account. You can get your free account and take control over your email by visiting the following URL. http://mailnull.com/