Why no namespace for a state?

Forgive me if this has been covered before, or if I'm wrong about this, but it seems to me that there is no way of specifying variables with a scope that covers all events in a state, but not other states. If I'm wrong, please tell me how, as I keep wanting to do this.

So we're either forced into using global variables (ugh, I hated those two words before I even met LSL) or variables within an event (or lower); there is no way of localising variables that span all events in the current state.

If there a good reason for not implementing this? After all, I certainly find that some data is used at that level. It seems to make sense.

Yep. You get locals and you get globals - nothing in between.


Cory Linden originally wrote the LSL language over lunch one day. There's a few things that are a little crusty on it. Don't get me started about the lack of switch/case..

Ah, that's a shame. But thanks for the confirmation.
Yes, while I agree with you completely, let's not go down the LSL-bashing road yet again!

at a guess, to be simpler to users not familiar with programming, or possibly because states aren't meant to control data, but event types...

although some of the choices for what events/functions will cross states makes that confusing, as timers cross, even set in another state, but listens don't, presumably dataserver/http type event/function pairs will, though there is little need or use, I'm not sure about target events, or sensor repeats...

I blieve any asynchonus event/function pair will cross states, and a few exceptions, like timers, are tied to the global level, while most others are tied to the state as you'd expect in a model that uses states to control available events...

granted it was supposedly written over a weekend (or was it a week, I forget?)