Clicking Cancel on a running Recipe causes hang 1 out of 3 times

Case number:845829-2002444
Opened by:g_b
Opened on:Wednesday, May 11, 2016 - 18:11
Last modified:Thursday, May 12, 2016 - 20:26

Clicking Cancel on a running Recipe causes hang (spinning cursor and says application is not responding and I have to force quit) 1 out of 3 times. This has been happening on several Macs for several years through 2 or 3 different OS upgrades. It happens with any Recipe that has been running for several hours. What do I need to supply to find this problem? Doesn't produce a crash log so may have to do a Sample Process with Activity Monitor when spinning. Don't know if it is the same problem but after client has been running for a day or so it makes the whole machine become very sluggish and have to be restarted. You can see the icons in Selection mode at bottom of screen being erased and then redrawn real slow. Maybe one icon will come back after 2 or 3 seconds and then the next one will be drawn and then 2 or 3 seconds for the next one etc.

(Wed, 05/11/2016 - 18:11  |  3 comments)

Joined: 09/24/2012
Groups: Go Science

This happens to me only when I load too many clients in parallel (for my Mac, 3-4 is usually a maximum).
It's mainly the RAM that causes this problem after many hours for big puzzles (it seems that the memory needed by foldit rises over time for each puzzle).
If I see that the mouse reacts too slowly, I close one of the clients.

spvincent's picture
User offline. Last seen 37 min 36 sec ago. Offline
Joined: 12/07/2007
Groups: Contenders

I'm not seeing this behaviour, although I only run one client on a machine.

Speculation: the offending scripts may not be using the Lua xpcall mechanism to gracefully handle errors.

frood66's picture
User offline. Last seen 3 hours 30 sec ago. Offline
Joined: 09/20/2011
Groups: Marvin's bunch

I've never seen the mouse react differently Bruno - but I see the slow down given time served by a client. Mac has historically had an issue with memory here. I take a fresh client regularly to try to combat this - mostly that works. It's not ideal (and it is a pain) but I've found no other solution.


Developed by: UW Center for Game Science, UW Institute for Protein Design, Northeastern University, Vanderbilt University Meiler Lab, UC Davis
Supported by: DARPA, NSF, NIH, HHMI, Amazon, Microsoft, Adobe, RosettaCommons