- changed status to resolved
fl.gaussian~ hang
I get an infinite hang when I connect a lone fl.frommax~ (that is being hit with a scheduler) to the left most inlet of fl.gaussian~. I'm assuming this is because its sending a zero length frame and this isn't allowed. I cant comment on whether thats a desirable behaviour but it makes it hard to patch together stuff as you have to have all your params setup above the fl.gaussian~ and then connect it. In an ideal situation it just doesn't compute if its getting a zero length frame and waits till you've hooked up a proper network. Sorry if this is an RTFM type error or a behaviour that is intended but if its not it might be worth addressing.
Comments (3)
-
repo owner -
Account Deleted reporter Sorry, Alex. I am aware of the distinction and went through and changed my terminology in the description but not the title. Ill compile and have some fun now
-
Account Deleted reporter - changed title to fl.gaussian~ hang
- Log in to comment
Fixed and pushed to GitHub - there might still be something to do here later to do with how inputs are read, but it won't hang.
Just for future reference - a crash is distinct from a hang. In a crash the application terminates - an infinite hang is not a crash. Sometimes you get a momentary hang then a crash, but that is different. Precision in terminology can help in diagnosing issues, although in this case it wasn't an issue.