[reactive] switching events = space leak?

[Don Vincenze]

Hi,

Is this list still alive ...

I am trying to learn the theory of FRP by reading the code of this
"reactive" library and Yampa. Not trying to get anything to actually
work here, analyzing implementations of the theory only serve to assess
the viability of it all. I.e. apart from bugs, is it stable, does it
scale?

For starters I'm worried about space leaks in switching events, e.g. see
the "Event of Events" that's the input to switchE. Those events are
lists that get produced by input producers but not consumed until
they're being switched into, is that right? So during this time of
producing but not consuming the list just grows = there is a space leak?

This is made a bit clearer in the book "The Haskell School of
Expression", chapter 15, which as I understand describes the
implementation of Yampa, not "reactive". However there isn't a
difference between the two in that respect as far as I can read the
code, or is there? Such as space leak would not just be a bug but would
need some design rework.

Don