Closed jproy closed 10 years ago
I've fixed the docs. However, the larger problem is that the "application pun" that happens in networks is just terrible--it always confuses me, and I came up with it. I think everything will read much much better if I replace
[out (foo bar baz)]
with
[out <- foo bar baz]
The difference is that the "call" to foo isn't actually a call, and foo isn't a procedure. Changing this should make it much clearer. Also, it will hopefully address the horrible hack that's currently in place where things like
[out (/ x 5)]
work, but for a really nasty reason.
No! Better idea! Dispense with the whole notion of signals as separate from functions. This means that initialization of signals has to become a call to a maker, but I think that'll actually read pretty nicely.
Arrrrrrr! No, the initialization doesn't quite work. Grr... Still thinking.
Reading "Signals & Systems" books, I thought that signals were functions time -> pressure and that systems were things that transformed signals… The doc will be important for people like me who have not this kind of culture :-) Macros are cool up to the point where too many things are hidden or just don’t "follow the rules" (e.g. function application). Sure you will converge to the best point of view. Thanks for your nice work and your time,
-jpr
Le 9 mai 2014 à 19:05, John Clements notifications@github.com a écrit :
No! Better idea! Dispense with the whole notion of signals as separate from functions. This means that initialization of signals has to become a call to a maker, but I think that'll actually read pretty nicely.
— Reply to this email directly or view it on GitHub.
Okay, I think I have a happy medium now. Should be fixed in 3280ea3
The example with control and sawtooth in the doc of lpf/dynamic does not work.