New Programmer

[bvisness]

Originally Posted by Greg McKaskle

I also learned LV long before there were shift registers and in many ways prefer them, but I now use both, typically for different things. I really like the compare to last with a shift register, but otherwise typically use shift registers.

Greg McKaskle

I almost never use feedback nodes, but I find myself using feed forward nodes quite frequently. They make more sense to me because the flow of the code keeps moving forward. If I'm in a loop, though, I'll always use shift registers.

[theboyer]

Here are my tele op and periodic vis neither has much extra really

[RyanN]

I looked at the code and I don't see anything wrong with it.

Basically, the starvation occurs if you don't call the Arcade Drive VI every 100ms. Having the VIs open (front panel or block diagram) can cause this. Try closing a few windows and see if that helps.

Seeing the starvation message is completely normal on non-built code (not saying it's good though). If you see it all the time after you build and deploy your code, then you have an issue.

Even with the message, can you drive the robot around? Do you notice it?

[Racer26]

Originally Posted by Greg McKaskle

I also learned LV long before there were shift registers and in many ways prefer them, but I now use both, typically for different things. I really like the compare to last with a shift register, but otherwise typically use shift registers.

Greg McKaskle

@Greg McKaskle: I really like several of the changes that have come about since the birth of the idea exchange a couple years ago. Things like shrinking the boolean constant, changes to local variable graphics, integrated labels for structures, better alignment of vi.lib vi's I/O and so on.

For my non-FRC applications, I REALLY like the JKI string-based state machine template.

[Ether]

Originally Posted by bvisness

I almost never use feedback nodes, but I find myself using feed forward nodes quite frequently.

Is "feedforward node" generally accepted terminology for a feedback node that simply has the arrow direction changed to make the wiring cleaner? If so, it seems unfortunate. The function of the node remains the same regardless of arrow direction.

[RyanN]

Originally Posted by Ether

Is "feedforward node" generally accepted terminology for a feedback node that simply has the arrow direction changed to make the wiring cleaner? If so, it seems unfortunate. The function of the node remains the same regardless of arrow direction.

As far as I know, they're only called feedback nodes. Changing the direction doesn't change their functionality. It just makes the code 'flow' more smoothly sometimes.

[bvisness]

Originally Posted by RyanN

As far as I know, they're only called feedback nodes. Changing the direction doesn't change their functionality. It just makes the code 'flow' more smoothly sometimes.

I guess it's not technically called a feed forward node, but that's what I have come to call them for reasons that are unclear to me. I think one of our mentors called it that, and it stuck.

But it does go forward...so I think I'm gonna stick to the "feed forward" name.

[Greg McKaskle]

Rather than feed forward, how about previous value or delay by one or some name that explains its purpose rather than its appearance.

Greg McKaskle

[Mark McLeod]

When I'm teaching it I use Greg's suggestion and just call it the "previous value"

[Ether]

How about calling it a Z^-1 node?

Just kidding. That doesn't help new learners.

I like Mark & Greg's "previous value" moniker. Students seem to understand that phrase immediately, and it's accurate.

@bvisness: the reason I said "feedforward node" is unfortunate is because feedforward has a completely different meaning in control theory.

[bvisness]

"Previous Value" is nice, so I think I'll try to call it that from now on. (And I was unaware that "feed forward" has another meaning in control theory...sorry about the confusion.)