[PD] nonlocal naming scheme

[Jonathan Wilkes]

Hello,
     I had another thread where I used my canvas "get" method to hide $0 in s/r pairs in an abstraction 

and make it possible to more easily specify the scope of a variable using two abstractions named [to] and [from].

Now I'm extending that concept to nonlocal signal objects: s~/r~ and throw~/catch~.  After looking 

through the list I've realized that some of the confusion about these objects occurs because the 

names are completely arbitrary and don't reflect the one-to-many/many-to-one division between them.*  

(Side question: are there any many-to-many nonlocal signal externals?  If so how does the cpu usage 

compare to throw~/catch~?)

So here are the names for my abstraction wrappers:

send~/receive~ = copy~/paste~

Thus with [copy~ foo] you are taking the incoming signal and copying it to whatever is at the 

receive-symbol $0-foo, which would be a [paste~ foo] object.  The one-to-many connection is that 

one can copy only one thing at a time (or on systems where you have multiple copy buffers they 

each have a different key combination or icon on the interface, similar to the way we give a different 

receive-symbol for each s~ in Pd), but you can paste it as many times as you wish.


throw~/catch~ = tobus~/frombus~

I think this one is self-explanatory.  However, I'm not sure if "bus" really implies many-to-one (as 

opposed to many-to-many).  Maybe there is a better metaphor.

Would love to hear if anyone has suggestions on these names.

Thanks,
Jonathan


* actually I think it's worse than that, because if you see that the control objects s/r are many-to-many, 

you could reasonably assume that s~/r~ would be many-to-many as well.