Capytalk ‘nil’

[Cristian Vogel]

Hi.

 

after 18 years using Kyma , I just tried this and it worked. Seems that the behaviour of nil acts like a sampleAndHold of the value computed by the true: expression

?freeze true: ( (?BackNForth + 1) * ?grainSize ) false: nil

[Cristian Vogel]

…or?    Maybe not. I can’t quite tell in the context I am working right now. Except that it compiles, so perhaps SSC can clarify 🙂

 

its not urgent

[SSC]

In the expression you propose, the nil case returns whatever is on the top of the stack, so the result is undefined (depends on what is happening in other parameter fields and Sounds).

[Alan Jackson]

Yikes. I’ve used the false: nil  expression a lot in the past kind of assuming it meant “if false don’t do anything”, but now I come to think about it the expression has to return something. It’s always worked fine when I’ve used it but I think I’ve been lucky!

I guess a way to use this reliably is to use an EventVariable, assign the new value to the EventVariable in the true: clause, put nil in the false: clause, put a comma after the expression then explicitly return the EventVariable.

Something like:

| ev |

ev := EventVariable new initialValue: 0.

(?freeze true: (ev <+ ( (?BackNForth + 1) * ?grainSize )) false: (nil)),
ev.

 

Is that the best way?

[Alan Jackson]

Ah, maybe you could simplify that a little by not popping the value off the stack when assigning it to the EventVariable like:

| ev |

ev := EventVariable new initialValue: 0.

?freeze true: (ev <~ ( (?BackNForth + 1) * ?grainSize )) false: (ev)

[SSC]

This last variation is the best. 🙂

[Author]

Posts