Btw, I had misunderstood ERN when, in 2018, I wrote the email message I linked in the Note section of my previous email.
ERN is an intuitive sense that some movement you are initiating is in some way a mistake. For example, quoting keys of a pair. It's often unconscious, and, aiui, ignoring it entrenches use of the mistaken action. Conversely, bringing it to consciousness strengthens the signal and the opportunity to "nip it in the bud" so to speak, sharply improving one's performance in the task at hand. That all said, I stand by the rest of what I wrote in that 2018 post, including the sense that folks' ERNs can be positively guided by careful PL design, especially if individuals' attitudes to WATs is a sense they are gifts (which is easiest to sustain if there really are indeed corresponding valuable gifts). And I also stand by what I wrote in my previous email to which this is a reply; I really do consider this named argument "wart" to be a great example of a gift that can be used to positively tune folks' ERNs. And this is why, in my first email, I tried to emphasize that from @Larry's perspective, the "warts" discussed in the original email were part of good design. Talking of which, I forgot to mention a detail of that good design that contrasts with something Joe said: > when you're accessing the hash value, the key *does* need > to be quoted (unlike in perl), As I noted before, one does not need to quote keys, because Raku has the very nice design detail of using `<...>` subscripts along with many other related uses of `<...>`. What I didn't note was how this avoided a "wart" in Perl that if one writes `foo` as a key of a hash subscript it is autoquoted even if one meant it not to be. I recall this being part of @Larry's rationale for the Raku design being as it is. -- raiph
