Why does squeak slowdown when several browsers overlap?

Klaus D. Witzel klaus.witzel at cobss.com
Sun May 21 19:03:49 UTC 2006


Hi Peace,

yes, I made the same observation (same in 3.8 and 3.9), simply did

  [13 timesRepeat: [FileList2 openMorphicViewInWorld]]

When you now move or resize a windows, you'll feel the gravity pull  
(excuse me, I meant the new formulation, the bending of spacetime ;-)

There shouldn't be such a big difference since the number of bits to be  
shuffeled does not change. But I also understand that these things are a  
matter of implementation (for example, one must find out *which* bits of a  
no longer obscured region must be displayed, per window). Not easy to make  
that perform better.

/Klaus


On Sun, 21 May 2006 19:58:59 +0200, Peace Jerome  
<peace_the_dreamer at yahoo.com> wrote:

> I am puzzling thru this question on mantis. See:
>
> http://bugs.impara.de/view.php?id=2084
>
> I recently came accross and interesting experiment
> when I tried to evaluate:
>
> "TimeProfileBrowser onBlock:
> [ 3 timesRepeat: [ Transcript open show: 100 factorial
> printString ]]."
>
> In a work space.
>
> This is a slight variation on the TimeProfileBrowser
> example on its swiki page.
>
> Running it, you can see the Transcripts updating
> bottom to top.  This suggest there is a cascade of
> changes that increases the rendering time by the
> square of the number of overlapping transcripts.
>
> Can someone with more skill with the tools than myself
> confirm this? And confirm whether this affects the
> overlapping browser situation as well as the above
> torture example?
>
> The payoff to this puzzle could be BIG. Its worth some
> community effort.
>
> Yours in service, -- Jerome Peace.
>
>
>
>
> __________________________________________________
> Do You Yahoo!?
> Tired of spam?  Yahoo! Mail has the best spam protection around
> http://mail.yahoo.com
>
>





More information about the Squeak-dev mailing list