On 14 May 2005, at 2:52 PM, Richard Gaskin wrote:

Compared to a field nearly anything else will be faster because of the overhead associated with all the other things fields have to do to display text in addition to storing it.

This topic has come up a few times lately, so I decided it was time to do some timing tests :-)

I had a field with 1000 lines of data put into a variable, then put the variable back into the field 50 times.
Then I did the same with a custom property, first getting the test data from the field (so it was identical) then setting & getting the property repeatedly.


The fastest method was using a custom property with messages locked.

Here are my test results:

using a field   740 ticks
using a field (with locked screen)      405 ticks
using a field from sub stack    230 ticks
using a field from a sub stack & locking the screen 235 ticks

The sub-stack's window was not open for these tests. Opening it made the times the same as for a field on the stack running the tests.

To get meaningful data for the custom property tests, I had to do 500 repeats, instead of just 50!
But dividing my average results then by 10, I got:


using a custom property 1.9 ticks
using a custom property with messages locked    0.3 ticks

This is almost 2500 times faster than the field method!!!!!

Removing the scroll bar from the field had almost no effect.

I really didn't expect to come up with such a big difference, but I expect it will only change my habits in certain circumstances. I like the convenience of being able to see my data when necessary so tend to use the closed sub-stack method. However if you have a vast amount of data to shuffle around and it needs to be done fast, this is a good method to remember.

Cheers,
Sarah

_______________________________________________
use-revolution mailing list
[email protected]
http://lists.runrev.com/mailman/listinfo/use-revolution

Reply via email to