Mac 3.0 benchmarks

John M McIntosh johnmci at smalltalkconsulting.com
Thu Feb 8 03:10:40 UTC 2001


>Eric Scharff <Eric.Scharff at Colorado.EDU> is widely believed to have written:
>
>>  The feeling that Squeak 3.0 feels sluggish (especially when compared to
>>  2.8) seems to be verified by the benchmarks.
>>  [snip]
>>  Are these slowdowns permanent?  Is there something I can do to avoid this
>>  over 10% slowdown?
>I see a similar slowdown on Acorn, and it seems to be almost entirely
>the cost of hitting the event polling prim(s) every 20mSec; change to a
>brain numbing 200mSec and get back most of the performance but loose all
>control over the pointer....
>
>It's just another aspect of the long running argument over pre-emptive
>multitasking versus co-operative; response versus raw performance etc.
>etc.
>
>In the particular cases of Mac & Acorn I suspect the polling takes a
>longtime precisely because it is a CMT call and many operations have to
>be done to get all the events.
>
>tim


Similar issue on the mac. The calls to get all the information are 
more expensive and you run the risk of giving up control to other 
application hence the poor numbers. In the past polling for mouse 
location for example used a simple/cheap call but this has been 
changed because of how the event driven stuff works.
-- 
--
===========================================================================
John M. McIntosh <johnmci at smalltalkconsulting.com> 1-800-477-2659
Corporate Smalltalk Consulting Ltd.  http://www.smalltalkconsulting.com
===========================================================================
Custom Macintosh programming & various Smalltalk dialects
PGP Key: DSS/Diff/46FC3BE6
Fingerprint=B22F 7D67 92B7 5D52 72D7  E94A EE69 2D21 46FC 3BE6
===========================================================================





More information about the Squeak-dev mailing list