Re: OODL: Java VM and bytecodes as models

[DeRobertis]

At 1:40 AM +0100 on 11/8/99, M. Uli Kusterer wrote:
>>I took a quick glance at it -- have not really tested it yet. But it does
>>have serious problems when you try to run the large 20,000 block test once,
>>then run it again. Took ~2 ticks/block. Had to nap on it :(
>
>Anthony,
>
> that's strange. On my slow Mac it always takes 0 (zero) ticks per block,
>which means it's always below 1 tick... hmmmm... back to the drawing board.

Hmmm... Are you testing the _second_ time of writing a lot of blocks? That
is, do it once, then immediatly do it again?

[I'll get to do more formal tests and examinations starting tomorrow --
nice long weekend].

>
> Did you try compiling it using MPW? Is that what the test came from, or
>will it get faster using MPW?

I have not recompiled it. Not sure how MPW will effect the timings.