gordonb wrote:My question concerns the amount of performance degradation eValid experiences when it is polling a DOM value? My understanding is that you can poll a specific element on a regular basis...but what is the performance implication of doing this?
-Gordon B.
As you may know, eValid is an actual browser and the overhead for most commands is less than 0.1%.
The eValid architecture was engineered to make the amout of system time that it takes to process a command as low as possible -- and in many cases the amount of time a command takes is not even measurable. The main reason for doing this was to make sure that tests for performance values, e.g. how long something takes to download or takes to sync, is 100% realistic.
In the case of a DOM polling loop, eValid limits the cycle time on such checks to a maximum of 100 Hz, that is, a 10 msec (10 millisecond) minimum interval. This value assures that measured sync times are within 1% of the actual value and further guarantees that eValid won't interfere significantly with the actual work being done in parallel by the browser while the synchronization polling loop is running.
Hope this answer helps.
eValid Expert Team
