The next situation has occurred a number of times in which wave~ shows abnormal behavior. An abstraction fills a buffer (loadbang) and produces, upon completion, from its outlet a set message with the buffername which is a symbol based on a #0 unique number. The outlet producing the set message is connected to wave~, which with a double click does show the content of the buffer, but a phasor~ connected to the inlet does not produce values on its outlet other then zeroes. Only after opening the abstraction and a manual click on the message, will wave~ respond to the phasor~ input. So, there exists a situation where wave~ is in limbo, on the one hand opening the correct buffer content with a double click, but not producing values on its outlet (showing that these two things are not necessarily connected). My impression is that this is so since the version of max for intel.
What I tried (change something and reopen the patch):
put a deferlow in between; no result
print the set message: result is that it works, until the print object is removed.
put a [zl reg] to capture the set message and bang it manually; no response (!)
change the order by sending to back, bringing to front; this apparently solves the issue once and for all.