[BUG] Button down event (2!=2 ?!?!)

[Ben]

Agreed!

I hqave to admit that I got a laugh out of watching a "2" and "2" go into an equal comparison and show up false.

Ben

[BJD1613]

Wiebe great find! Thanks for some interesting troubleshooting this morning. 🙂

[EdDickens]

I just checked this in 7.0 and 6.1 on WinXP and it's been this way since the event structure was released.

Something does need to be fixed here. What is being displayed and what is being operated on is not the same value. And that is just wrong. I guess NI will have to decide which is the correct value.

Ed

[Wiebe@CARYA]

"Ed Dickens" wrote in message
news:50650000000500000034CA0100-1079395200000@exchange.ni.com...
> I just checked this in 7.0 and 6.1 on WinXP and it's been this way
> since the event structure was released.
>
> Something does need to be fixed here. What is being displayed and what
> is being operated on is not the same value. And that is just wrong. I
> guess NI will have to decide which is the correct value.
>
> Ed

Try this (if you didn't already).. Display the value, but also convert the
value (e.g. to I32), and then display this value. The converted and
unconverted values are different.

Anyway, if (when) they fix it, what whould be prefereable?

AFAIK, the input should be what is pressed, and the wire that has this value
should
be the same everywhere. The value in this wire should be seperated
from what you put into (a completelly different, although both named button)
output.

Regards,

Wiebe.

[EdDickens]

Want to get confused even more?

Turn on the new "Show Buffer Allocations" tool in 7.1 after you add the numeric conversion.

New buffers are created for the original U16 indicator, and on the output of the number conversion. So it looks from this that the y ternimal on the equal function and the I32 conversion use the same buffer, but the original U16 indicator creates a new one even though no operation is done on the value. The new I32 indicator reuses the new buffer created by the conversion function.

So where (and why) is this new buffer created for the U16 indicator getting it's data from. It looks like the front panel is updated before any of the code is run. That's why the indicator is showing the value before it's bee
n modified by the filter event.

I hope some one at NI is reading all this.

Ed

[Philip_Courtois]

Finally, a real bug!

I�ve investigated the behavior of the Filter Events and it seems that the filtered �Button� terminal shares the memory space with the incoming �Button�. This means that when the �Equal?� function is executed, the �Button� parameter gets updated to 1.

Putting a sequence around the �Equal?� function will allocate a new memory space at the tunnel, so the filtered �Button� will not refer to the same memory, hence the equation will work.

A stable workaround for every Filter Event case is to always force the filtered terminals to be updated as the last thing.

Anyway, I will report this bug to R&D for future improvements.


Thanks to everyone for troubleshooting!

Best regards,
Philip C.
Applications Engineer
National I
nstruments

[Ben]

"Finally, a real bug!"

Would you like to look at my collection of CAR#'s?

They go back to LV 5.1 (or was that BV 2.1?)

My favorite was the "wires from nowhere" bug in LV 6.0 and 6.0.1. I believe this one forced LV 6.0.2.

Let me know and I will be more than happy to pass all of the bugs I find to you.

Ben

[Philip_Courtois]

Hi Ben,

I know you have been an invaluable help for the LabVIEW world when it comes to troubleshooting and answering questions - thank you for that.

If you are aware of any bugs that haven't been reported to the LabVIEW R&D team, please make sure to let us know. Especially if you can reproduce it in the most current version of LabVIEW.

Send us an email using our Technical Support site. Then we will be able to improve LabVIEW for future versions.

We appreciate feedback from anyone, thanks.

Best regards,
Philip C.
Applications Engineer
National Instruments

[Ben]

I will do that!

Unfortunately I do not have new ones to report.

Give me time.

Ben

[Ben]

Hi Philip,

Take a look at this thread.

http://exchange.ni.com/servlet/ProcessRequest?RHIVEID=101&RPAGEID=137&HOID=506500000005000000CACE0100&HTHREAD=000118474&UCATEGORY_0=_49_%24_6_&UCATEGORY_S=0

Is this a bug?

Ben