On Wed, 2 Apr 2003, Brian Brown wrote:

> I see from one of Aaron's posts what class to look for so I can try some
> Squeak ones... the only image I've played with is the tinySqueak one,
> but that is an older release right?

TinySqueak.image by Kevin is a 3.0-based image.

> I recently put Aaron's DynaPad image on there from the website (on an
> AudioVox with a 400Mhz XScale, 12MB of ram for Squeak). I'm seeing *at
> least* a full second delay before any menus pop-up in tinySqueak,
> Dynapad seems more responsive for some reason.

The Dynapad R0.1 image is based on Squeak 2.8, while TinySqueak is
based on 3.0.  While not the source of all of the speed difference, a lot
of it is likely because of the differences between 2.8 and 3.0.  There has
been some talk on this list about the generally change in responsiveness
between Squeak version 2.8 and 3.0.  Some people seem to think that this
is absurd, but a lot of people have noticed that there is indeed a
difference in how fast Squeak feels that occured somewhere in the series
from 2.8 to 3.0.

> I can do benchmarks on each one if folks would like, if someone could
> point me to a decent image we could use to benchmark?

If you're just using TinyBenchmarks (print-it "0 tinyBenchmarks"), just
about any image would be fine.  TinySqueak.image would be a fine image for
testing this.  The version of Squeak, especially in version between 2.6
and 3.6, doesn't really effect the bytecodes and message sends per second.
VM does, but not the image.  So as long as you're working with a PDA with
at least 14 MB or so of real RAM available for Squeak such that it will
not have to use a memory mapped swap file, TinySqueak would be fine.  The
differences in how different images feel speed-wise isn't due to the kind
of performance these benchmarks measure.

I've thought about and done a little work on a very small image (=< 1 MB)
that just runs some benchmarks and gives you the option of emailing them,
printing them in the transcript, or writing them to a file. If more people
are interested, I could polish this up and share.

Regards,
Aaron

  Aaron Reichow  ::  UMD ACM Pres  ::  http://www.d.umn.edu/~reic0024/
  "if i don't stay true to live and hate, how do i differentiate
              between chasing cream and chasing dreams"  :: atmosphere