The User-Defined Config Element doesn't work as you describe in my instance. In my instance, functions didn't work in it either, just like the Test Plan (before I fixed the Test Plan). Now, I've fixed it just like the Test Plan and now both behave the same - ie functions are resolved once at test start.
Question is, should I check this code in before the 2.1 release? I judge it somewhat risky code. -Mike On Wed, 2005-08-17 at 16:04 -0400, Michael Stover wrote: > Funny you should mention that as I've just reworked the TestPlan element > so that functions can be used in it's user-defined variables section > (they will be evaluated once - useful for, say ,the __P function). I'm > wondering whether I dare risk committing it for the 2.1 release. > > The bug you mention regarding the config element is very interesting. > Sounds like a neat feature actually :-) I'll take a look.. > > > -Mike > > On Wed, 2005-08-17 at 19:37 +0000, Peter Bernier wrote: > > Hello, > > > > I'm just trying to figure out whether or not this is a bug. > > > > When you have a User-Defined Variables Config Element at the Test-Plan > > level (as opposed to on the Test Plan node itself), variables are > > re-calculated with every request (as opposed to at the Thread Group or > > below level where they're calculated once). > > > > This is different from the behaviour of this Config Element anywhere else > > in a Test Plan. I'm just wondering whether this is expected behaviour, > > or not? > > > > Specifically what's got me is that in the Component Reference (Sec > > 15.4.14) It says "The User Defined Variables lets you define variables > > for use in other test elements, just as in the Test Plan.", but this > > isn't quite accurate since you can use functions (ie, __Random) on the > > UDV Config Element whereas you can't on the Test Plan's UDV. > > > > The reason I bring this up is that I've recently switched to storing > > TestPlan-wide variables in a UDV Config Element at the Test-Plan level > > as opposed to on the Test Plan node itself. (due to Bug 35872) > > > > - Peter Bernier > > > > --------------------------------------------------------------------- > > To unsubscribe, e-mail: [EMAIL PROTECTED] > > For additional commands, e-mail: [EMAIL PROTECTED] > > > > --------------------------------------------------------------------- > To unsubscribe, e-mail: [EMAIL PROTECTED] > For additional commands, e-mail: [EMAIL PROTECTED] --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
