Require stability annotations on fields of tuple variants

[petrochenkov]

This wasn't done in #29083 because attributes weren't parsed on fields of tuple variant back then.

r? @alexcrichton

[alexcrichton]

Oh dear those attributes, that's quite the line to give to the compiler! Maybe we should just stop stabilizing enums :P

@bors: r+ 8ea7b88

Thanks!

[petrochenkov]

Hopefully it will be better in 2017, after the next snapshot

[bors]

⌛ Testing commit 8ea7b88 with merge e51661b...

[bors]

☀️ Test successful - auto-linux-32-nopt-t, auto-linux-32-opt, auto-linux-64-debug-opt, auto-linux-64-nopt-t, auto-linux-64-opt, auto-linux-64-x-android-t, auto-linux-cross-opt, auto-linux-musl-64-opt, auto-mac-32-opt, auto-mac-64-nopt-t, auto-mac-64-opt, auto-win-gnu-32-nopt-t, auto-win-gnu-32-opt, auto-win-gnu-64-nopt-t, auto-win-gnu-64-opt, auto-win-msvc-32-opt, auto-win-msvc-64-opt

[durka]

How about fields of struct-like variants?

[petrochenkov]

How about fields of struct-like variants?

They required annotations already.

[durka]

I figured, but I didn't get any errors when I forgot to include them in
#30884.

On Fri, Mar 4, 2016 at 7:18 PM, Vadim Petrochenkov <notifications@github.com

wrote:

How about fields of struct-like variants?

They required annotations already.

—
Reply to this email directly or view it on GitHub
#30898 (comment).

[durka]

@petrochenkov see durka@003120a (but it compiled before that commit too)

[petrochenkov]

@petrochenkov see durka/rust@003120a (but it compiled before that commit too)

It compiled because unstable annotations are automatically inherited from parent items.

[durka]

Oh, I see. Makes sense.

On Fri, Mar 4, 2016 at 7:43 PM, Vadim Petrochenkov <notifications@github.com

wrote:

@petrochenkov https://github.com/petrochenkov see durka/rust@003120a
durka@003120a (but it compiled before
that commit too)

It compiled because unstable annotations are automatically inherited from
parent items.

—
Reply to this email directly or view it on GitHub
#30898 (comment).

[petrochenkov]

Unfortunately, annotations on variant fields are not checked properly yet, so the fields in durka@003120a may be insta-stable (AFAIK, it's the first occurrence of named variant fields in the staged API).

[durka]

How could you name them without using the unstable variants and the unstable enum itself?

[petrochenkov]

You probably couldn't. This is a concern for (hypothetical) stable variants with unstable fields.