Say I've got 4 U8s that I want to join into a U32 - right now, I have to build 2 of the U8s into a U16, then build the other U8s into a U16, then build the 2 U16s into a U32. I'd like to see these primatives growable to each datatype width.
If it should be "linear", should be a '0' pre- or suc-ceded?
Norbert
Norbert ---------------------------------------------------------------------------------------------------- CEO: What exactly is stopping us from doing this? Expert: Geometry Marketing Manager: Just ignore it.
I'd say make it linear and fill with 0 in the MSBs.
There are only two ways to tell somebody thanks: Kudos and Marked Solutions Unofficial Forum Rules and Guidelines "Not that we are sufficient in ourselves to claim anything as coming from us, but our sufficiency is from God" - 2 Corinthians 3:5
+1 for growth by 2^N - let the application programmer explicitly wire the 0's in the high or low bytes as a self-documentation exercise and diagram readability.
I could use this in many places. I would prefer a linear expansion where any padding can be defined to be undefined or the type default. For my use cases, undefined would work fine as I would only pull the data back out or use the bits that I explicitly put in. Kudos!
What would happen if you pulled out an undefined element? Would you expect the VI to break? Or would you just expect the developer be savvy enough to remember what's supposed to be in there?
The
I wouldn't expect the VI to be able to break since the Join and Split may be in different VIs (i.e. this is probably something we wouldn't include in the type of the wire). As a compiler implementor, I would hope users would expect ANY value of the given type to possibly come of the node, although the most likely case is that a default-default for that type would be given. Yes, choosing that puts extra burden on the developer to ensure they handle it properly, similarly to how you would use typecast or any other reinterpret-like functionality.
