04-30-2014 12:25 PM
In my earlier attempts in doing precision arithmetic with Time Stamps I neglected to check for overflow of the fractional seconds.
Being off by a full second is certainly unacceptable!
Addition or subtraction of Time Stamps can be done with full precision if we cast each of them to a cluster of I64 and FXP<+,64,0>
then unbundle and do the math separately. If the fractional result is outside the range of [0,1] then it wraps, overflow is sensed
and the I64 result is adjusted by ±1 second. Then the results are bundled back together and cast to a Time Stamp.
Hopefully this is the last update... if anyone is interested, I have two VIs, one to add, the other to subtract, two Time Stamps with full precision.
Let me know by posting to this thread.
Thanks,
Jon
03-13-2017 06:01 AM
I discovered a reference to CVI in Bug ID 236193 in LabVIEW Known issues documents for versions 2016, 2015, 2014.
That ID is not present in CVI Known Issues documents (searched in 2009, 2010 and 2015)
03-14-2017 01:43 PM
Thank you for the heads up, I have create a CAR for the documentation with CAR ID 636203. This forum hasn't been posted on in a while, so feel free to create a new thread for issues like this in the future.
03-14-2017 06:16 PM
Well, I posted here following a link for suggestions on the known issue page: others could post here too, I think this page should be considered active until the link points here.