06-25-2018 12:46 PM
@etvg wrote:
There isn't really a purpose,
That's what I thought. One potential "band-aid purpose" is the path control. If it is not in the sequence as shown, it might get read before the operator has a chance to enter the right value. However if this is an issue, it might point to other potential architectural problems instead. 😉
Hard to tell without seeing the entire code...
(Often sequence structures are overused, limiting the compiler into a straight-jacket of forced serialization where things could be done in parallel otherwise.)
06-25-2018 12:50 PM
If you want to look at the code I believe attached it to my second post on this thread, it starts with LIdA3. Thanks for the tip though, I'm relatively new to LabVIEW so everything that will save me headache in the future is valuable information.
06-25-2018 01:19 PM
I am not sure that I have time to look at the full code at the moment, but it seems highly unsafe to accumulate all data in a tunnel and write everything at the end. If the computer would crash (power failure, OS issue, etc.) or the file IO would fail (e.g. accidentally select an location that is not writeable, or some other illegal path entry) all data stored in the autoindexing output tunnel is permanently lost. It would be significantly safer to open the file and write the header once before the loop and then append data as it arrives inside the loop. (or use a consumer/producer system if the loop speed is critical). Only close the file once the loop has finished. Currently, you could even run into memory issues because the data in the tunnel grows without limits if the while loop would never stop for some reason.Having large and growing data structures is never efficient.
Just some thinking points. 😄
06-25-2018 01:21 PM
Thanks for the pointer!