Closed equilet closed 7 years ago
reassigning to Richard as David hasn't responded
Worth mentioning that the other day I was going over some Github details with Ed... It turned out Ed hadn't validated his email with Github yet, nor were notifications ala @EdmundCampion
demarcation or assignment working for him.
This directly relates to issue #19 and may be solved shortly.
I've gone ahead and added a 64-bit perform routine for cambio~. I tested to make sure it operates as expected across both 64/32 bit launches of Max. We'll need to build this on Windows whenever we release next, but it's currently in the repository and added to the XCode project, /src
folder, etc.
added to windows makefile
It is useful, and looking through the C code, easy to maintain.
Because this is a small issue, I'm assigning to you for the time being. If this takes more than 20 minutes, please assign to Adrian.
Case reference: http://cnmat.berkeley.edu/casetracker/case/13112 http://cnmat.berkeley.edu/casetracker/case/13154
Lubow to Campion: "This is actually something to bring up with John - cambio~ is not the same as [snapshot~ 1] -> [change 0.]
In my estimation, to get this functionality at the bounds of the signal vector, you really need to be doing what cambio~ does. John, let me know if I’m wrong about this.
I think that we should continue to support cambio~ (even if it ends up being called something else), as it’s very useful for issuing control-rate changes that are bound to the signal vector w/r/t timing. It is also hanging on to the last few values for comparisons to detect the change with. All of this is expensive outside of the context of the perform routine.
It’s a matter of adding it back into the externals release, which shouldn’t be too hard, even if we want to rename it."