Interface ground to a halt

Wilhelm Sanke sanke at hrz.uni-kassel.de
Tue Apr 13 17:27:17 EDT 2004


On Mon, 12 Apr 2004, Norman Winn <norman at mrsystems.co.uk> wrote:

> I am actually getting somewhere. I managed to create a large bunch of
> buttons and get them cycling through the random colours.
>
> Fortified with my success I decided to script the creation of an array
> of rectangles, 62x25 of them.
>
> This took rather a long time.
>
> I have not yet worked out how to set the 'left' of my rectangles at
> creation time so each row just got piled up on the same spot.  So, once
> they had all been created, I tried to 'select all' with a view to
> lining them up edge to edge (this is a great feature).


Here is a script to create a matrix of rectangles - edge to edge and not 
overlapping for a 40 X 40 matrix:

on mouseUp
   put 100 into toright
   put 100 into vert
   lock screen
   lock messages
   put 40 into tline
   set the width of the templategraphic to 9
   set the height of the templategraphic to 9
   set the filled of the templategraphic to true
   set the linesize of the templategraphic to 0
   set the backcolor of the templategraphic to green # or whatever
   repeat with i = 0 to 39
     add 9 to vert
     put 100 into toright
     repeat with j  = 1 to 40
       add 9 to toright
       put "Test" & (i * tline + j) into tname
       create graphic tname
       set the topleft of graphic tname to toright,vert
     end repeat
   end repeat
   choose browse tool
 end mouseUp

> Unfortunately when I did 'select all' everything ground to a halt for  
> many minutes. Am I overloading things by doing this? Maybe rectangles
> use more resources than buttons?


Revolution has sometimes serious problems to handle larger numbers of 
controls in the IDE (These problems do not exist with the Metacard IDE). 
Compare my post "(long) Transparent IDE elements and other problems" of 
March 25 to this list.

This is especially true when the controls overlap as in your case. Monte 
Goulding discovered this when he tried to recreate one of my tests 
stacks and was working to improve the new Standalone Builder.

> One problem is that I don't get the beach ball (OS X) to indicate
> things are going on. The whole app just becomes inactive - but not
> locked or crashed,


In some cases I have had waiting times of up to 45 minutes!

I think (or better "hope") that the Rev team will address these problems 
soon.

Regards,

Wilhelm Sanke




More information about the use-livecode mailing list