LabVIEW
MAX tasks vs generated tasks - opinions
So generally I use MAX. I setup the channels the way I like with the scales and devices the way I want. Then my application will create a test panel based on the signals already made in MAX, and my software uses those virtual channel names.
The reason for this in my mind, is because it is something that can be changed without having to build a new version of software. If a channel on a DAQ card got burned up, but we just need to continue testing, we can re-write the signal, change the signal in MAX and then my software doesn't need a config change, or a new build. Another benefit is being able to change the signal scale, if a new sensor is used.
I also have my software on startup try to see if all the virtual channels are there. If they aren't then it programatically imports the NCE file. This usually means I need to manually look at the config to see that the import happen successfully but it works well for most setups. Just some times if you have two DAQ cards of the same model you need some checks to ensure the virtual channels went to the right one.
That' being said I have worked on distributed systems, where tasks had to be created at run-time. I just like relying on MAX for channel setup if I can. It means less work for me.
