RFC: Make struct optional in mutable struct

[ararslan]

After using this for a while, I've decided that I find mutable struct to be quite long, and I find it odd that it doesn't match struct in being a single word. Thus this PR makes struct optional. Mutable types may now simply be defined as mutable X end rather than mutable struct X end, though the latter still works. Once upon a time, type was optional in immutable; we had immutable and immutable type. This is analogous to that.

I imagine this may be controversial, but hopefully it can at least be considered.

[fredrikekre]

RIP comment: #20418 (comment)

[JeffBezanson]

Sorry, but I'm 👎 here. (1) I like the presence of a noun; the basic idea of type and immutable type was good, it just used the wrong words and the wrong default :). (2) In isolation this syntax would be ok, but this requires adding a reserved word, which I don't think is worth it.

[ararslan]

Mostly I just dislike the multiple-word definitions, so I figured having this one be optional may be a decent compromise. Granted, I don't particularly like the idea of anything being optional like this; I just miss concisely defining types.

[kmsquire]

@mutable macro?

[KristofferC]

I don't think defining structs is common enough for a super terse syntax to be needed.