posted 16 November 2014 18:50            

I completely understand why a uniquely named piece of memory needs to be used but when a sound containing uniquely named memory is encapsulated and then duplicated (in a timeline or another sound) errors occur because of conflicts with named memory duplicates.

This ruins the benefits of encapsulation because to fix it you need to 'unencapsulate' the sound then track down and fix the duplicate names which is often a time consuming process that results in a much more complex sound.

It would be great if the encapsulation process could somehow manage named memory such that multiple copies of an encapsulated sound could be used easily.

posted 17 November 2014 12:48            

All the best
Chris

p.s.: I am not in front of my Kyma... but when i remember correctly for delay-lines "Private" or private... does the job.
Of course this does not make sense for a memory write because we want to "read" from it somewhere in our sounds.

posted 17 November 2014 16:18            

Very usefull for encapsulation of sounds with MemoryWriters is the ParameterTransformer Prototype. It allows you to access the boolean parameters of a MemoryWriter, Sample Prototype and others with a ?greenVariable (silent, global, cyclic, reverse etc.)

for example:
snd silent isNil ifFalse: [snd silent: snd ?silent].

Somehow i remember that they have to be lowercase in the ParameterTransformer script, although they are named with Uppercase names in the prototype description.

While you create the "new class from example" give a corresponding boolean value as a default -> true or false. In the class editor you declare it as a boolean.
And in the interface it will appear as the typical field with a cross to toggle.