ID caching: how to test?

[Peter Haworth]

On Sat, May 11, 2013 at 11:50 AM, Richard Gaskin <ambassador at fourthworld.com
> wrote:

> Any tips on how to test this, anyone?


Maybe I'm oversimplifying, but I executed this script in 5.5.4 and 6.0
(apologies in advance for any asterisks inserted into the script by gMail)

*on* mouseUp

   *local* tStart

   *put* the milliseconds into tstart

   *repeat* 10000 times

      *get* the short name of card id 1002 of stack "lcstackbrowser"

   *end* *repeat*

   *answer* the milliseconds - tstart

*end* mouseUp

In 6.0, it took consistently around 380ms.  On 5.5.4, it
took consistently around 280 milliseconds.  So unless I'm misunderstanding
what's supposed to happen, it takes longer with the caching in 6.0 than it
does without in 5.5.4.  Which makes me thing my feeble brain doesn't
understand it either!

Pete
lcSQL Software <http://www.lcsql.com>