Programming for the future

[evarsanyi]

I realize this doesn't make any sense short term, but...

 

It would be much more flexible if the programming facilities could be used to directly build state machines which could hold internal state (ie: variables) and were passed events directly.

 

All the current functionality could be implemented in terms of these. It would allow one to write code at the level the actual network runs. You don't know that unit X is 'on', you get an event saying it 'went on' (or maybe you don't even though it did); you don't know that the time is XX:XX right now (by definition that's false), you do know that XX:XX has passed from the future to the past.

 

I'm very happy with the functionality provided in 2.5! I have found that implementing solutions in a real world means understanding the ugliness of the real world and trying to wallpaper over it means either you lose flexibility or things can't be made to work reliably.

 

Again, really happy with this stuff, just some input for the next spin.

[Mark Sanctuary]

Variables are coming; I think it was in the next 2.6 release.

[Chris Jahn]

At some point in the future we'll be adding variables and other features. Variables in particular would be useful in creating a state machine.

 

All the current functionality could be implemented in terms of these. It would allow one to write code at the level the actual network runs. You don't know that unit X is 'on', you get an event saying it 'went on' (or maybe you don't even though it did); you don't know that the time is XX:XX right now (by definition that's false), you do know that XX:XX has passed from the future to the past.

I'm not sure what you are getting at here; the current programming model is entirely event driven.

[Mark Sanctuary]

The PLM modem language is too low level and cryptic for average users to program in. The ISY programming model is about as low level you would want to go.