Work In progress (WIP) #8.5

[stefanhendriks]

I have got word from Greg that there are a few other bugs on memory side, causing SEGV faults. I will fix them.

The mass suicide bug is being worked on, i know what is wrong. I don't know yet how to fix it

[Gismo]

For the moment, I have replaced the #8.5 with the #8MM. I have deleted all RBN RBX files and started again from scratch. So far, the HLDS has been running for 19 hours or so. The reason for the switch backwards was that the HLDS was closing every 8 hours or so. I have re-enabled error reporting to see if that can tell me where the fault lies but as the only thing that was changed was the RB MM. The control group I set up has performed well (killed a lot of the clients lol) even down to learning the matrix mod which had the bots really confused before! The only problem I have now is that the old WIP gives me that SZ_Getspace error which I had attributed to Statsme (2.7.1) but after shutting down statsme for 4 hours it was still there.

Can't wait for the next release as RB just gets better and better.

Keep up the great work Stefan!

Try the latest relaese of statsme.

[Gismo]

Think you missed the point there - it was not Statsme that caused the problem (Statsme was off when it shutdown again - HLDS had caused etc), but the WIP8.5. The latest version of statsme is not fully compatable with CS1.5 at the moment anyway. I know this as I installed it without RTFM! lol. I reinstalled 2.7.1 from the backups same as I reinstalled the WIP8 which has been running for around 24 hours now. The memory leak problem is still present ((partially)fixed in 8.5)but I am using FreeRAM from Bysoft to free the memory blockage. It also seems that other people are having similar problems the their HLDS crashing - possibly from the other memory problems Stefan has discussed with a de-bug post. I just thought that Stefan would like to know thats all.

And if I wanted to talk about Statsme I'd just go to their forums!