Herb[Romulus2] 2003-05-28 10:43:03 | Just curious, have you ever watched the size of the *.sav files during a simulation. Mine currently start around 2.2MB and get as far as 3.8MB during the simulation. The bigger it grows, the better the final result. Could that be used to abort a current simulation, if the growth rate isn't sufficient for a good result? ------------------------------- I'd say more, but I can't reach the keyboard from the floor. |
David 2003-05-28 11:54:30 | It would be more direct to use the commandline client and guillotine poor performing simulations based on the particles remaining data you can get from that. |
[DPC]Stephan202 2003-05-29 02:04:02 | But then you'd have to watch the screen all the time. I think what the TS wants, is to create a script/program that runs at the background and checks the file size of the *.sav file every X minutes. When the file is not growing fast enough, the client is shut down and restarted. --- Dutch Power Cow. MOOH! |
Herb[Romulus2] 2003-05-30 00:36:58 | Yup, that's righ, the client could have some kind of a performance meter inside, that discards a lower level design whenever detected. That would save a lot of wasted calculation time. I'm watching several clients over a network, it could be also a feature to restart the client with a new data selection, by just remotely temporary presence of a file (RC5 clients used such a feature, but only to turn it completely off). ------------------------------- I'd say more, but I can't reach the keyboard from the floor. |