Inefficient code - Solutions (loopless)

[BNig]

Victoras,

very nice. 
I went that road too, I did a srcOrReverse and than I wanted to test the
maskdata for non transparency (would reduce the amount of data by 8: not
comparing 2 pictures and then just working on one byte for each pixel. Did
not get the Alpha channel working. Would have been interesting for large
images.

If you are interested: in your 2 images there are 25920 pixels that are
different and it took me 90 milliseconds to find out. Thanks to Bill's
algorithm. 
(your stack works on a Mac, I am using one)

regards
Bernd 




viktoras d. wrote:
> 
> here is one more simple, loop-less, dirty but fast script (130 millisec 
> on 1.5 GHz celeron) which produces image of differences. Will not work 
> on MacOSX because it uses srcXOR ink :-). I imagine more variations and 
> combinations are possible also using bitXOR to process imagedata in a
> loop.
> 
> http://ekoinf.net/revolution/imageDif.rev
> 

-- 
View this message in context: http://www.nabble.com/Inefficient-code-tp24226458p24288892.html
Sent from the Revolution - User mailing list archive at Nabble.com.