[grumpy mode] The servers ran like crap last night. I Played three or four maps and saw too much rubber banding, red plugs and general hiccups going on. Maybe the FPS was better at times but what is the point if it drops to zero at other times. It's not just the fault of NS2 itself (Luajit, garbage collector stuff) I played on Iron Horse's server and it ran smooth as glass all the way through. Perhaps the player limits or plugin selections or hardware choices are the cause, who knows but a tic rate of 13 is fail. fix it. [/grumpy mode]
Seems to be better the last two nights. Thanks.
Hey TinCan. Grumpy! This doesn't sound like you.. Read this http://steamcommunity.com/groups/dmd...16813435628558 The servers run like crap whenever there is an error in the ns2 source code. When an error is detected, a trace is started to find the root cause of the error. The error loops indefinately until map change. At which time the VM is destroyed and a new instance is created. As the errors are caught and logged - along with the trace and exact cause - I then pass that data to the developers. Currently there are I think three errors still left on their "to do" list. The servers run perfectly fine - up until an error is triggered. One error as an example is if someone stars evolving after a myst has been deployed. Getting a second myst on the same egg caused script errors that would repeat about 15-18,000 times (till map change). This and many other issues have been fixed in build 209 which we're running now. The hardware is sound, and if tickrate drops, that's due to an error in the source. nothing more. nothing less. Here is a perfect example for you. http://www.diamondgamers.net/aaarust...luajit_2_1.txt This is build 308. This is with tracers running (to take the error and find the cause) You can see that there were over 20,000 traces ran. and at the start of it - it shows the error. and the end of it - shows the data that's needed to fix it. This error was when a marine picked up two catalyst packs at the same time. Boom - tickrate drop as the engine kept running the source over and over and over again - failing each time ( as it was doing what it was told to do! ). This ran until the map changed. This error has been fixed in build 309. I know it's frustrating - try being in my shoes and handling everything - it's a lot of work to say the least. But know that I am working with the devs to assist them with data collection etc. The reason why you had no issues on IronHorses server was he didn't have an error come up which looped endlessly - as he's running a Windows server. If you see my original steam group post - it explains how the devs tested the build on windows but NOT linux. Hope that helps.
Yea, I was a little salty when I made that post. I just want the game to be great you know, and yea it was the looping that got to me. Like I say its running better now. Originally Posted by Rusty he's running a Windows server. His dad said it was a Linux box so I'm not sure on that. If NS2 does run better on Windows, well, then I think the fix is obvious.
Forum Rules
