@tehn @Prnts et al
SUPERCOLLIDER FAIL tends to mean that the sclang process has entirely failed to launch for some reason. using the maiden REPL won’t help in this case because there is nothing on the other end of the websocket. (with duplicate engines you get far enough that the sclang interpreter works, but can’t compile its class library.)
the most common cause of launch failure is that there is already a sclang process holding the requested UDP port. it’s not too hard to create this situation by using/abusing the debug scripts; rebooting usually fixes, but there are ways to bork it worse by playing with the jack environment or with systemd.
as @okyeron said, another other possible reason is a general problem with systemwide network management, though this should not just happen after an update.
finally, it is possible to have engine or core norns SC code that totally crashes the sclang process on init, one way or another.
in all cases, starting sclang directly from the shell is the best diagnostic step and will generally yield immediate clarification.