ID caching: how to test?

J. Landman Gay jacque at hyperactivesw.com
Sat May 11 14:27:24 CDT 2013


On 5/11/13 1:50 PM, Richard Gaskin wrote:

>  > The image caching just happens automatically. Benchmarking would
>  > involve timing a procedure in 5.5.4 and then repeating it in 6.0.1 to
>  > see the difference.
>
> Right, but using what script?  Which references cache and which don't,
> and which operations should show the benefit of caching?

I read it to mean all images, imported or referenced, and all object 
references ("card ID xxx", "btn ID xxxx of card 3", etc.) are cached. So 
to test, I guess one way would be to run some image-heavy stack that 
uses duplicate images on different cards. The old way was to reload 
every image when the card was first drawn. The new way is to keep images 
in a cache in case they're needed again. So make a stack and place the 
same image on multiple cards. Navigate through the cards. Try it in both 
5.5.4 and 6.0.1.

They also say caching works for referenced images, so do the same thing 
with those instead of importing the images into the stack.

Another test might be to load an image dynamically by script. Store a 
couple of images somewhere, and run a repeat loop that loads the text of 
an image object alternately between the two. In an uncached scheme it 
should reload each image from scratch on each call. In a cached scheme, 
the two images should be ready to display instantly.

I think.

-- 
Jacqueline Landman Gay         |     jacque at hyperactivesw.com
HyperActive Software           |     http://www.hyperactivesw.com



More information about the use-livecode mailing list