Remove let expressions support

[p-bakker]

Looks like the let implementation supports let expressions, something from ES4 that never made it.

See:

• https://web.archive.org/web/20150516161726/http://wiki.ecmascript.org/doku.php?id=strawman:let_expressions
• http://www-lia.deis.unibo.it/materiale/JS/developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/let.html#Non-standard_let_extensions
• https://bugzilla.mozilla.org/show_bug.cgi?id=1023609

OK with tossing it out?

[tonygermano]

General advice tends to be to not remove things that work, because you never know who might be using it. I could see excluding it from language version 200+.

[gbrail]

Right now, our compatibility practice has been to not break things for language level < 200, but for language level >= 200 we should try to match the spec. So I would be scared to remove this unless we can only do it for the latest language levels. I realize that other JS engines don't do this -- they just keep moving forward. But again with Rhino embedded into so many things I hope that by keeping it compatible we can encourage people to upgrade and take advantage of the other good things that we're doing.

…

On Fri, Jun 4, 2021 at 4:23 PM tonygermano ***@***.***> wrote: General advice tends to be to not remove things that work, because you never know who might be using it. I could see excluding it from language version 200+. — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#923 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAD7I25VRQMNQBX3SCJH6BDTRFN5VANCNFSM46CNIYUQ> .

[p-bakker]

Think it's doable to just disable this when the language version is >= 200, but...

The reason I came across this was that I was looking to see what it would take to make the const (more) standard compliant, by basically make it use the same logic as the let implementation, but then I saw that the let implementation had all this extra code for a one-time draft feature that never made it into the spec (but I'm sure it can be achieved with some conditions here and there)

But it does being up the question how to do a standard compliant const implementation or any fix for that matter that corrects non-standard behavior, with a behavior change as a result. Should as such fixes be tied to a specific language feature?

If you look at #326 and it's linked cases, fixes for all of them would constitute behavior changes, from non-standard, to standard-compliant behavior

[p-bakker]

Also see @rbri's comment: #326 (comment)