Mutate Sidechains feature takes too long to stop.

Case number:845813-991306
Topic:Game: Tools
Opened by:BootsMcGraw
Status:Open
Type:Bug
Opened on:Wednesday, December 14, 2011 - 04:01
Last modified:Friday, January 25, 2013 - 03:27

I have a relatively powerful computer, yet when I attempt to stop the Mutate Sidechains function, it doesn't stop for at least forty-five seconds. I have one that is still running, over three minutes after I hit the stop button. I'm going to have to --- once again --- close the game and re-start it. Not efficient or fun.

Can someone please program the game to understand that "stop" means "stop now", and not "stop one of these days, when you feel like it, if you feel like it"?

(Wed, 12/14/2011 - 04:01  |  26 comments)


Joined: 04/19/2009

Completely agree!

Joined: 10/11/2011
Groups: None

Yes I can have the same issue.
I have also noticed that sometimes shake and wiggle cant remember what stop is either

beta_helix's picture
User offline. Last seen 3 days 4 hours ago. Offline
Joined: 05/09/2008
Groups: None

assigning...

CFC's picture
User offline. Last seen 3 years 16 weeks ago. Offline
Joined: 11/03/2010
Groups: Contenders

As far as I know, mutate will only stop at the end of a full cycle, (and shake too). You shouldnt need to close the client - it will stop, so long as you believe in The Force.

CFC

Joined: 06/17/2010

Yes, "stopping" message is sometimes annoying ;]
Shake and mutate need a handbrake when we press "stop" button.

Joined: 12/06/2008
Groups: Contenders
Status: Open » Closed

I don't seem to notice this problem as being as severe as it was six months ago, so I am considering the problem resolved. I'll re-open a ticket if I see it, again.

spmm's picture
User offline. Last seen 36 weeks 3 days ago. Offline
Joined: 08/05/2010
Groups: Void Crushers
Status: Closed » Open

I have noticed this during and post CASPX on all mutate puzzles using a manual mutate. Devprev selection, Win7.

infjamc's picture
User offline. Last seen 2 years 40 weeks ago. Offline
Joined: 02/20/2009
Groups: Contenders

Hmmm, maybe this game could use an "emergency restart button" (which would close the game and open another instance) as a stop-gap measure? (It's good for getting out of other types of hangs anyway.)

spmm's picture
User offline. Last seen 36 weeks 3 days ago. Offline
Joined: 08/05/2010
Groups: Void Crushers

bump - any scripts are taking waaaaay to long to mutate as well

Joined: 09/25/2012
Groups: None

Is anyone still observing that "Mutate Sidechains feature takes too long to stop"?

If yes, can you point me to a puzzle - better yet, the name of a solution file - that reproduces the problem particularly well? (It would be great if you could also provide your computer hardware and operating system info.)

I'd like to look into the issue further and see if I can improve the situation.

Thanks!

Joined: 12/06/2008
Groups: Contenders

"...can you point me to a puzzle - better yet, the name of a solution file - that reproduces the problem particularly well?..."

Every. Last. One.

Joined: 06/17/2010

Shake and mutate are affected on ANY puzzle.
More sidechains selected - it takes looooonger.
"Stopping" message can last up minute or more....

marie_s's picture
User offline. Last seen 1 year 45 weeks ago. Offline
Joined: 05/18/2008
Groups: None

Better in devprev on symmetric hub.

beta_helix's picture
User offline. Last seen 3 days 4 hours ago. Offline
Joined: 05/09/2008
Groups: None

Do you mean that the problem is better on this puzzle, or this puzzle is an even better example of how bad it can get?

If (by some miracle) the problem isn't as bad on this puzzle using devprev, how about on any other design puzzles (in devprev vs. main). For example 636 or 638?

It would be great if the latest devprev update fixed this issue!

marie_s's picture
User offline. Last seen 1 year 45 weeks ago. Offline
Joined: 05/18/2008
Groups: None

For me,in dev prev, the time to stop mutate is higher with local selection than with no local selection.
This time is higher than to stop wiggle, is shorter than before and not annoying.

Joined: 08/24/2011

bump. the problem is really annoying on 656. could this be linked to a sort of memleak? now foldit is using like a third of my ram (4Go), which will drop hard if i start anew. i know the waiting will drop significantly too if i reboot foldit and then will increase by small steps until it becomes unusable and my ram reaches its fullest. i've never seen such problems with shake.

Joined: 05/19/2009
Groups: Contenders

I too experience this slowness in puzzle 656 on an i7-920 2.66 GHz running 4 clients (using 4 Gb and 8 Gb free) with Win 7 64-bit latest 2012-11-19-52243M-win_x86-devprev.

663 Mb after running for a while, stopped and restarted with exactly the same position, now uses 531.

I also recall I reported several revisions of devprev ago there STILL was a memory leak after the mega-leak some time ago. To my knowledge that was never fixed.

So here you have a CONFIRMED memory leak. Upping priority.

Joined: 05/19/2009
Groups: Contenders

After just ONE action of 'mutate all', the M key, it leaves me with 585 Meg, a staggering 54 Megabyte increase... After another wiggle or two and a new mutate we're at 604...

tamirh's picture
User offline. Last seen 6 years 13 weeks ago. Offline
Joined: 05/11/2012

If you close the puzzle and open it up again are you still at that same level? Whenever you first use a feature like mutate it will use up some memory. When it's done running, the memory isn't necessarily cleaned up immediately.

Joined: 08/24/2011

from my experience, closing the puzzle has no significant change. it can hasten up the thing a bit but gets to "waiting mode" very fast.

Joined: 08/24/2011

bump. almost solved on normal design puzzles, but it's a pain on symmetry ones.

katfish's picture
User offline. Last seen 4 years 49 weeks ago. Offline
Joined: 01/09/2013
Status: Open » Open

Thanks for bumping, everyone. Looking into it presently.

Joined: 11/10/2007
Groups: Window Group

Reassigning.

Joined: 11/10/2007
Groups: Window Group

We've found an issue that would cause it to take longer to stop on symmetry puzzles, so we'll get out a fix for that as well.

Joined: 12/06/2008
Groups: Contenders

The halting of the mutate function is definitely longer for symmetry puzzles. Although I haven't timed it, I believe that the more mutable residues on a protein, the longer it takes the mutate function to stop.

Thank you for looking into this, Seth.

Sitemap

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