On Wed, 16 Aug 2023 09:19:37 +0200, Ovid <curtis.poe@gmail.com> wrote: > 1. Can type failures be downgraded to warnings or disabled entirely? When you're working with a multi-million line >code base, having production come crashing to a halt because your "integer" recieved 3.14 instead of 3 is going to >get types ripped out quickly. Thus, this approach would likely be of little use for existing systems. while i have other issues with the OP proposal (such as them uploading a broken dist to cpan and grabbing toplevel perms without discussion) i have to ask about this question if i declare a moo/se attr as isa => Int, and send in 3.14, moo/se will also die to my knowledge the only way around that is eval, so why would downgrades be needed, rather than proper testing and error catching? or is there a way to downgrade the type errors in moo/se? -- With regards, Christian WaldeThread Previous | Thread Next