Hi all, thanks for the constructive feedback. On Fri, 29 Jan 2021 at 15:52, Marco Pivetta <ocram...@gmail.com> wrote:
> > Gonna be the usual person saying: let's not clutter functions with more > behavior than what's needed :-) > > If you need to validate a hash for being "well formed" rather than > "matching", then write a separate function dedicated to that, rather than > increasing the complexity of a pre-existing symbol. > > Marco Pivetta I would definitely vote for that, too. It feels to me like throwing an exception on invalid algo / malformed hash should be the default behaviour, and that the use case brought up by Anthony, although very relevant, should be the one that should test for validity using another function (or pass an extra parameter). That being said, it's indeed a BC break. Would it be small enough to be allowed to hit 8.1? I don't know. — Benjamin
