Support 18 clients

[madwort]

Might attempt to do an 18-way session for LIO first Sunday of April - would need to up the patching limit from 11 in order to do that!

[sandreae]

I'm pretty sure all the connecting logic is ready for dealing with more clients, it's generating the ladspa plugins where we need some automation instead of a hard coded list. I had a brief go here using a numpy function to spread positions (based on number of clients) over a range. It could be a good way to go, most of the changes were in get_panning_positions which is looking smart.

jacktrip_pypatcher/ladspa_plugins.py

Lines 15 to 33 in 75e98dc

	def get_panning_positions(self, number_of_clients):
	if number_of_clients == 0:
	return []
	if number_of_clients == 1:
	return [0]
	if number_of_clients == 2:
	# Just map points between -0.5 and 0.5 for shallow 2 client panning
	return list(np.linspace(-0.5, 0.5, number_of_clients))
	if number_of_clients == 3:
	# Fixes the strange case of 3 clients
	panning_positions = list(np.linspace(-0.5, 0.5, number_of_clients))
	panning_positions + (panning_positions[0] - 0.01)
	panning_positions + (panning_positions[2] - 0.01)
	return panning_positions
	return list(np.linspace(-1, 1, number_of_clients))

Problem here is that the positions aren't ordered in any way, so some of out routing logic is probably (definitely) broken. I'm not sure if that only effects the special 1 / 2 / 3 client situations though. More thought and testing needed 👍

[madwort]

Problem here is that the positions aren't ordered in any way

not sure I follow what you mean here. In the current version, they're ordered by how close they are to 0, we could replicate that with e.g.

>>> sorted(np.linspace(-1, 1, 15), key=lambda x: abs(x))
[0.0,                 0.1428571428571428, -0.1428571428571429, 0.2857142857142856, -0.2857142857142858, 
 0.4285714285714284, -0.4285714285714286, -0.5714285714285714, 0.5714285714285714, 0.7142857142857142, 
 -0.7142857142857143, 0.857142857142857, -0.8571428571428572, -1.0,                1.0]

(compare to unsorted)

>>> np.linspace(-1, 1, 15)
array([-1.        , -0.85714286, -0.71428571, -0.57142857, -0.42857143,
       -0.28571429, -0.14285714,  0.        ,  0.14285714,  0.28571429,
        0.42857143,  0.57142857,  0.71428571,  0.85714286,  1.        ])

and then the current logic might not have to change toooooo much?!

[madwort]

or could do a deeper refactor, of course!

[sandreae]

Problem here is that the positions aren't ordered in any way

Haha, yeh, what did I even mean there....? They're ordered, well, in order. So I meant we'd need to do what you suggested 👍

[sandreae]

or could do a deeper refactor, of course!

What do you have in mind? 😏