HI!

Quick version:

My folder structure:
/mother-patch.pd ( main pure data patch first loaded and running... )
/patches/1/main.pd
/patches/2/main.pd
/patches/3/main.pd
...

How i could open /patches/1/main.pd triggered by some action in the mother-patch.pd?
How i could close /patches/1/main.pd triggered by some action in the mother-patch.pd?

Long read

I'm attempting to clone the Critter and Guitari organelle ( link ).
This instrument is basically a computer running libpd and running very cool pure data patches
You can read more about my project here ( link )

I have a mother pure data patch that it's first loaded when the device is on, this patch is doing some [send] and [receive] operations related to the knobs/keyboard/volume/led and it should be also managing the loading ( opening and closing ) of the child pure data patches ( mentioned above ). This child patches are receiving the actions from the mother patch.

If I open the mother patch and the child patches manually, everything works fine. But now I need to OPEN this CHILD PURE DATA patches with a object within the mother patch.

I've been testing [open] , for opening the main.pd of the child patches but it does not work.
I've been testing [pd] , for opening the main.pd of the child patches but it does not work.
I wouldn't like to modify the original organelle patches.
I wouldn't like to end up loading all the patches using [pd ...] and inserting a [switch] object inside every main.pd file in the patches in order to enable only the dsp of the desired patch.

It would be perfect If i could have a folder with all the patches and load them within the mother patch with some kind of object. And I would also want to be able to CLOSE the pure data patch and open another ( changing patches... )

Every little and big help woul be MUCH APRECIATED!

THANKS!!!!!!!

@jfernandezriverola Why don't you want to use [switch~]?

If you really want to open/close patches as needed, you're going to interrupt the audio engine every time you open or close a patch.

@jancsika thanks for the quick answer!

Several reasons:

• Using [switch~] implies to slightly modify every (third party) patch that I want to use. I'd prefer to use them as is, and simply put them on a folder.
• As i understood, if load let's say 20 patches with switch to off, it'll consume less cpu but will still be consuming some anyway. Is that right?
• I don't mind to interrupt the audio engine every time I want to open / close patch . The open and patch close is an analogy for changing a "Preset" on a synthesizer, so it's ok I won't be emitting audio in between preset changes.
• Keep in mind that the mother patch will alway be loaded, i want to open close only the "child pure data patches" ( ore 'presets' ).

What do you think?

THANKS!

I'm experimenting right now with: [openpatch] / [opa] but failing anyway...
It sounds like exactly the thing I need but I can't make it work.

My folder structure:

My pure data code:

And still does not open..

@jfernandezriverola If you just want to automate the process of opening and closing patches, there are ways to do it. There's a message you can send Pd to open a patch. There's also a message you can send to a canvas (or abstraction) to close it.

I can't remember the exact selectors atm, hopefully someone else can fill them in...

@jfernandezriverola It's in the Pd "manuals" folder...........
2.open_close.pd
David.

@whale-av That's it!

@jancsika said:

@whale-av

@whale-av and @jancsika thank you very much.

The problem I was having with pd open is that I was trying to make it work with the relative path.
I see in the docs you show that "usually you don't get it to work with relative paths".. and that's right, once you try it with ABSOLUTE path it's working nicely.

THANK YOU BOTH for the quick answers, what an awesome forum and community.

If someone is interested, I ended up with this in order to change patches.
I'm forcing the quantity of patches to 8, but in the future I'll try to force the length to the number of patches in the folder ( with the [folder_list] object ).

It's a very simplistic approach, but it's working fine!

@jfernandezriverola

Hi, I need a patch that loads automatically a list of child patches (30/40). Each pacth after a given duration time get close and the next patch get open. Maybe I can solve with this patch, but I can't make it works. I get the error:
"trigger: i: bad type"
on the [t i b b] object.
Any help?
Thanks!
Matteo

@deframmentazione-geometrica Hello......
it looks as though it should be an "f"
Triggers will send known Pure Data message "types" in the "right to left" order from their outlets..... and also convert them to "bangs" where there is a "b".
The types are:-
a ..... anything (useful if you are not sure)
b ..... bang
f ...... a float (number)
l ....... a list....... i.e. list 1 2 3....... or woof 4 5 6
s ...... a symbol ...... i.e. woof
p ..... a pointer ...... to a scalar in a list
so the message you see is saying that trigger cannot use/ does not know what "i" means....... and so there is an error in the previous post....
David.

@whale-av

Hi David, thank you for this clear and useful answer.
Matteo

I am accomplishing something similar, in a different way, so I don't have to use full paths. Chek out pdpi. In MAIN.pd I dynamically load a patch like this:

When it receives a float message on programchange, it dynamically loads the patch, then sends a synthetic loadbang message (dynamic loaded stuff doesn't trigger loadbang.)

To wire things up, I have a catch~ L, catch~ R hooked to dac~ and s control, and s note hooked to ctlin and notein

After what it has been said in this topic maybe I'm going to ask obvious things, I'm not a progammer so be patient.

Well, I want the children patchs open randomly through a first order markov chain. I found a patch with a markov chain matrix that works well changing simple frequencies mother markov 2.pd . Now, how can I write the children patchs list in the message to load the patchs? I did some attemps but I got the simultaneous opening of patchs or the crash of pd. Any suggestion? Thanks

@deframmentazione-geometrica Hello again! I am not completely understanding, and I don't have [markov~] on my system so I have had to guess the meanings of the messages. Is this anywhere close to what you might need?
mother_markov_2.pd
I don't understand what you mean by "children" in your patch, as you are not opening any "sub-patches" that I can see......... so I have just built a small system to bang in the messages randomly and after a random amount of time....
I hope it makes sense......
David.

@whale-av Hi David, thanks, it makes sense... Now I can open randomly the sub-patch and this is already nice.
Going a step further and following your patch I did some clumsy attemp to make the markov matrix "drives" the choise of the next sub-patch to open http://www.algorithmiccomposer.com/2010/05/algorithmic-composition-tutorial-markov.html.
without anything coming of it mother markov 2.1.pd
Matteo

@deframmentazione-geometrica Hello Matteo...... what drops out of the outlet(s?) of the [markov~] object?
Put a [print] object and connect it to the outlet to see what it is sending. Look at the terminal to see the output of the [print] object...... and that will give you some clues about how to trigger your patches...
David.

@whale-av Hi David, in the end I changed my mind along the way and I did this mother markov 3.pd, using the most common and understandable method I found on the web using random and moses objects. At moment I abondoned the idea to make it with markov object and transition matrix.

Thanks for your help.
Matteo