Io boxes losing values in subpatches

hello yous

(i don’t actually believe this is a bug, since someone else must have come across it, but i didn’t find anything about it on the forums… so my question would be: what the hell is wrong with me?)

i discover this weird behavior where io boxes in subpatches won’t keep their values. it goes like this:

new patch
make a subpatch
duplicate that subpatch
in the duplicated subpatch, create an io box (any kind), connect it to a second one, change its value, cut the rope. value in the second io box will switch back to its former state.

fun fact:
it only happens when the first subpatch is hidden.

is that some expected behavior? am i missing something fundamental (or maybe just a forum thread)? or isn’t that plain weird?


madame vee you are quite alright, nothing wrong with you. i can reproduce the same behaviour (when i hide the first of the subpatches) and i don’t think it is intended.

thanks for the confirmation. always good to know ur not crazy…
anyone else? devs? it can get really annoying at times…

easy workaround would be to not hide your root?

stumbled over this thread describing a super weird behavior and figured out it is actually not plain weird.

even if you have multiple instances of your patch, its description only exists once. if you move a node in one patch instance it will move also in the other one… and in the end (after naming it) it will be stored once on your harddrive, right?
so your “patch description”, your “patch source code”, your “patch class” only exists once. an input that is not connected to anything obviously can only have one value it will later on store on the harddrive. a value that is the same for all your patch instances.

now you could argue about which of the values (of the many instances) that were flowing on the link just a second ago shall be copied into the now disconnected input. it could be the value of the first instance or just the value of the instance where you actually disconnected the link.

but the one way or the other. only one value across all instances can be stored in the now disconnected input. it is actually quite good that this happens right away and not delayed after storing and reloading the patch.