[racket-dev] [racket] Stuttering problems with big-bang

From: Robby Findler (robby at eecs.northwestern.edu)
Date: Thu Aug 8 10:31:27 EDT 2013

Sorry: first post. Here's the link:

https://gist.github.com/Universalist235/6171371

Robby


On Thu, Aug 8, 2013 at 7:47 AM, Robby Findler
<robby at eecs.northwestern.edu>wrote:

> I don't think so no. But we could try.
>
> His code is linked in the dirt post in the thread.
>
>
> On Thursday, August 8, 2013, Matthias Felleisen wrote:
>
>>
>> Truth in advertising:
>>
>> big-bang skips frames in certain circumstances -- when handlers take too
>> long -- so if the GC kicks in when a handler runs, it is possible that the
>> eventual result is a skip.
>>
>> Robby: do you think Jack's program would benefit from a forced gc before
>> big-bang runs? (I can't find Jack's code.)
>>
>> -- Matthias
>>
>>
>>
>>
>>
>> On Aug 8, 2013, at 6:40 AM, Robby Findler wrote:
>>
>> Hi Jack: the entire system freezes when the GC happens and the numbers
>> you report match up with what I'd expect to be very annoying stuttering.
>> (And just for the record, it doesn't skip frames -- it just doesn't update
>> anything or draw any new frames during the GC.)
>>
>> Can you tell us more about your machine? What OS and what processor?
>>
>> Robby
>>
>>
>> On Wed, Aug 7, 2013 at 9:55 PM, Jack Firth <the.game.creator6 at gmail.com>wrote:
>>
>>> I started up the program from the command line with the garbage
>>> collection debugging information, and saw some interesting stuff once the
>>> program was up and running. Every second or so I'd see a very very slight
>>> stutter, which would be followed by a garbage collection lasting either
>>> 16ms or 15ms, slightly longer stutters would report as GCs of 31ms, and
>>> long freezing stutters would be around 109ms. I think perhaps big-bang is
>>> doing something where it's skipping a frame or two as a result of some sort
>>> of garbage collection. All the on-tick etc. events are performed for the
>>> frame, it just doesn't seem to be drawn. I can't fathom what's going on
>>> here.
>>>
>>> _________________________
>>  Racket Developers list:
>>  http://lists.racket-lang.org/dev
>>
>>
>>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/dev/archive/attachments/20130808/ebeae048/attachment.html>

Posted on the dev mailing list.