stim buffer is resource intensive

[GoogleCodeExporter]

Using open loop stim (and therefore closed loop stim) on 3.16 ghz, intel core 2 
duo machine running windows xp (3.25 gb ram) causes the machine to run behind 
on recording.  Task manager shows that both cores are maxed out when this 
occurs.  NR isn't supposed to be restricted to super computers, so we need to 
figure out a way to make this more efficient. 

possible solutions:

1. could switch from loop to event driven algorithm for loading new buffers 
(similar to the way that reads are handled)
2. could switch from 'update every sample' to 'update when sample differs from 
previous' algorithm
3. could add configuration tool to set different processes to different 
priority levels based on user needs/hardware capabilities (ie, if user has 4 
cores, could have option to set the stimbuffer background worker to real time, 
the closed loop background worker to real time)

Original issue reported on code.google.com by RZellerT...@gmail.com on 13 Dec 2010 at 11:16

[GoogleCodeExporter]

this isn't fixed by any means, but substantial improvements have been made by 
switching to the event driven scheme (I forget which revision this was)

Original comment by RZellerT...@gmail.com on 25 Aug 2011 at 7:47

• Changed state: Started

[GoogleCodeExporter]

This issue is fixed in 0.7.0.0

Original comment by jonathan...@gmail.com on 2 Nov 2011 at 1:59

• Changed state: Fixed