Hi,

I’m trying to [clone] abstractions which already use $1, $2, … for internal send/receive pairs and more. Is there any way to avoid “pushing” argument order in [clone] while still addressing individual instances (-x doesn’t do the job)?

Of course I could replace $1 by $2 and so on inside my abstractions, but it’d be great to keep one version working in and outside of [clone].

Maybe there’s an obvious workaround I’m not seeing, hope this makes sense.
Thanks!

@Gabriel-Lecup I think you will have to shift your arguments up by one..... but I am not sure that will help you because extra arguments applied to clone will be the same for every instance which will break your patch (they will all be talking to each other through any send/receives).
You will need to use $0 for internal messaging, or the clone 1st argument.
Those are the only options.
The clone 1st argument you can use for external send/receives.

If you need other arguments that are different for each instance (symbols or floats) you will have to push them in with "this" and "next" messages and route them to value stores [symbol] or [f] within the abstraction, or selectors [sel] or [route] as necessary.
David.

Many thanks David as always for your quick and insightful reply!

I do not need to comunicate between instances (thanks for the this and next tip though), neither send / receive locally within each instance. The question is if I really need to shift arguments up a notch from non-cloned abstractions: you answered that very clearly.

I guess I need to have a clonable
version and a clone-free version for concerned abstractions ...

Thanks again!

One more thing: it seems I can avoid the problem by dynamically creating instances of the abstractions I need to clone ... Is there any reason to use clone rather than dynamic patching?

@Gabriel-Lecup No, I don't think so.
[clone] is great for simple single argument abstraction cloning.
Copy/paste was the way to do that previously, and then you had to change all the arguments manually.
If you needed thousands of copies it was a real pita.
Dynamic patching with counters was the solution.
Dynamic patching is also great when you need a variably constructed patch, and will allow you to have many different arguments and automate the build process.
I have a mixer patch that I use for work, where I set the number of inputs and outputs and the whole mixer is dynamically patched to that size with all the metering and control windows.
Dynamic patching is a very powerful tool in Pd.
David.

Great, thanks for the insight.
I'll follow that guide then: [clone] for simple single argument patching, and dynamic when I need more control. I'm quite new to dynamic patching, but starting to get my head around it, using a lot of the [text] object too ...
If it's fine with you, I'll be sharing the toolbox I'm working on soon, and will ask for your feedback? Very grateful for your help on all this.
Best,

Hi @whale-av,

I'm trying to understand how clone works, so I found this post.

I would like to control one parameter of my cloned synth separately for each instances, for example to have different note's volumes.
I tried to use 'this' and 'next' messages. So I create 'this $1 parameter $2', then in my abstraction I recuperate the signal through the first inlet and I route it to volume. I did the same with 'next'.
But it doesn't work, I guess it works in a different way.

What I'm doing wrong?

Thanks you

@JamesStick You can send messages to a specific clone (through an inlet) by adding the clone number before the message........ so just leave off the "this" "next" or "all" from your message..

.......... example.zip

Don't forget that clone numbering ($1) starts at 0 unless you use the -s switch to force numbering starting at 1 or some other number....... [clone -s 1 module 4 1 0] for the screenshot below will create clones 1-4 instead of clones 0-3.

Ignore the other stuff in the example unless you need to send messages between identical clones.
So [3 data( sent into an inlet will only be received by the clone which has the value 3 assigned to $1.
The 3 is stripped as the message arrives...... so clone 3 receives "data" on the inlet.
You could also send the message in "wirelessly" of course using a [receive whatever-$1] within the cloned abstraction.
David.

@whale-av

Thank you very much, it works!

I'm happy to learn how to use clone.

Bests