On Fri, Sep 25, 2026 at 01:33:43PM +0200, Étienne Pflieger wrote: > Oh man, how stupid I feel now… This was also the advice of Peter Bex sooner > (sorry to not have react to it before :/). I was pretty sure to have tested > that sooner in my iterations, and I remember receiving an error such as "xxxx > is not an output-port" or something like that, and nothing when using the > opposite.
Some procedures do check their argument port's direction (in/out), others merely check that the argument is a port, seemingly for no good reason (the types.db entries *are* correctly annotated with the port's direction). I will send a patch to chicken-hackers to improve this, because it's definitely a bit of a footgun. Thanks for pointing this out! Cheers, Peter
