Is TestStand 3.5 slower than TestStand 3.1?

[adi@4]

I ran an example in both TestStand 3.1 and TestStand 3.5 without any process model. I found TestStand 3.5 slower when compared to 3.1. With tracing enabled it was approx. 10 sec slower and with tracing disabled it was approx. 7 sec slower.

I have attached the example.

[MathieuSteiner]

Hello Adi@4,

I do not reproduce this behavior here. I cannot see the differences - understand that I do not notice differences when I count in my head

How do you test this ?

Regards,

[adi@4]

I measured the time taken to execute the test. For small programs it does not matter. But for huge programs, the time difference may be quite noticable.

[AllenP]

Are your station options/report options/etc... identical?  I can understand this happening if you are using On-The-Fly reporting in 3.5 and regular reporting in 3.1.  I will try to reproduce this.  Also, are your search directories identical?  A large amount of search directories could slow down your application, especially if it has a network path very early.

Allen P.
NI

Message Edited by AllenP on 02-14-2007 09:26 AM

[AllenP]

I was able to run your sequences in both TestStand 3.5 and TestStand 3.1 and recorded the time differences.  I was not able to see the time differences you describe.  Most likely, you have some setting that is different between your versions.  You can test this by deleting the TestExec.ini and StationOptions.ini files in both versions of TestStand (or backing them up), and running it again.  I also disabled result collection in both versions of TestStand (but also benchmarked with it turned on in both).

The entire sequence you sent me was able to run in under 2.5s in both 3.5 and 3.1.  I was running without using a model, tracing disabled, and no result collection.  This leads me to believe that you have some settings difference between the two versions.

Allen P.
NI

[adi@4]

Yes. I had some additional tool settings in TS 3.5.

When I removed them, performance is now same in both versions.