CW3 v. 2.0 on OS X

[dmtheorist]

It appears that the 2.0 version of CW3 has begun to run extremely inefficiently on my MacBook Air (13" Mid-2013; 1.7 GHz Intel Core i7)...it is currently consuming ~87% of CPU while running (per Activity Monitor).  I didn't notice this heavy resource use under previous versions, and a day or two ago CW3 crashed completely.  I think I've found the log of that incident (see attached); any advice would be greatly appreciated!

EDIT: For what it's worth, I downgraded to v. 1.82 and played the same map I was playing when it crashed before.  It runs fine under v. 1.82, while only consuming ~20-25% of CPU

[knucracker]

If you have a save of the map you are playing where you observe a difference in performance between 1.82 and 2.00, attach it here. That would be very useful for understanding the issue. Most of the changes between 1.82 and 2.00 are steam related and for the non steam version they aren't even compiled in.

[dmtheorist]

Here is the save game I was playing when it crashed.  I have also re-installed v. 2.00 to see if I can duplicate the crash and send a crash report.  I'll keep you posted.

EDIT: Trying again, with attachment.

[dmtheorist]

UPDATE: After re-installing (i.e., unzipping) version 2.00 I have been unable to duplicate the performance issues I previously experienced.  Maybe something went wrong with the unzip when I did it the first time (?).

I'll re-post if I encounter any problems again.

[knucracker]

That does happen more often than you would think.  I have several reports where simply reinstalling the game on OSX (unzipping it again) fixes issues.  Most of the time the issue is a crash, but I suppose a race condition might also be a problem.  Could have been a problem that caused an internal exception to get thrown that was writing to the log file rapidly (or something like that).  That would cause a slow down...

[ak]

I've had all recent versions of CW3 running at ~90% processor even in the start menu on my mid-2009 MacBook Pro.  Came here wondering if there was any solution but it seems not.

[knucracker]

Don't measure performance on the start menu, but rather in a game you have started and paused.  The start menu doesn't have a frame rate limit in place so depending on you GPU and driver it can appear higher than the actual  game (counter intuitive, I know).  When an actual mission starts there is a 30fps limit in place and when the game is paused there is very little logic being calculated each frame.  So use that as a baseline for your normal CPU usage then see what happens when you unpause the game.

Also note that if you use the steam version the steam overlay can cause some issues with some GPUs/drivers and performance.  You can disable that to see if it makes a difference.

Lastly, test on a small simple mission (like the first few story missions) to establish an idea of baseline performance then move out from there.  Some custom maps can have lots and lots of complex scripts that use as much or more CPU time as the rest of the game.