I reported this behavior to the QlikView support during last august, the error occured because of a macro executed at document opening and a timing issue (which explains why it happens randomly). The bug was corrected in SR1, so maybe the bug was reintroduced into SR2 or maybe your error is a different one.
The ID assigned to the bug was #23058.
We encountered the exact same issue/error. I submitted a ticket to Support but because we were running a patch error and had no way of testing it in the unpatched SR2 (7320) they closed the case. The fact that you are getting this error helps us with our case to them.
We don't have any macros and we pin pointed it down to it only happens when you try to select something in a table on one of our sheets. There really is nothing special about this table chart - just that it uses stoplights but we have other that use that and this doesn't happen.
Hope this kind of helps!
The "Pure virtual function call" error is not one specific error, but a generic one that indicates that a method on a former morphed object is called when the object it self is being destructed, and the base class does not support the call.
This error can show up in more than one place without being the same error.
This is classified as a bug that should be debugged properly by a Support Technican or anyone with knowledge on how to catch exceptions in a debugger. Report them to email@example.com. I know that we caught one of these little buggers just a couple of days ago in SR2.
I've had ideas on doing a blog post on debugging Qlikview, but so far the complexity and "gotcha's" that's painfully apparent in debugging software scenarios has stopped the evolution of this. But if people are interested, it might be worth putting some time into.
We reported this issue to Support to but they told us that becuase we were running a customer patch they couldn't help us until we upgraded to the next service release. They did ask if we could reproduce it the regular SR2 but we didn't have a way to test that out.
Was there something else i should have done? Is this something that would be fixed in SR3?
Any information you could provide me would be wonderful.
Customer patches don't go through the same tests as Service Releases, and therefore are not supported on the same level. So that is probably the answer, yes.
As I said before; this is not a specific error, and I don't know the details in the fix of the last one, but if you can reproduce an error it should be reported. If they match other reported issues they will be consolidated at support level.
But basically, and error like this can happen in the Qlikview IE Plug-in (OCX) and will most probably crash Internet Explorer. Steps to reproduce are important, if they can be recorded.
We have the same problem, but we are working with V9 R4, the same version on server and plug-in.... this error appears when the user select with double-click some record in a pivot table. The pivot table has some columns with semaphores and is possible to do a fast-change to a sumarized table.
My customer can replicate the error each time, when they repeat this steps.
Then the bug is still present on R4.... do you have some news from technical support about this error??
Regards and thanks a lot!
Gabriela, I would recommend you to contact support at firstname.lastname@example.org (or via any channel that suits you) to open a new case on this, since this is probably not we can find a workaround for. It might be data dependant, but if you want the sam structure as you have now, we probably need to troubleshoot and debug the issue at support level.