Spec text inconsistency

[caitp]

From IRC:

[16:18:12] <caitp> rwaldron: do you notice some inconsistency between
    http://rwaldron.github.io/exponentiation-operator/#PostfixExpression
    and http://rwaldron.github.io/exponentiation-operator/#sec-update-expressions ? 
[16:18:40] <caitp> it looks like the "diff with last spec snapshot" wasn't updated  
[16:19:07] <caitp> it's sort of confusing to point that out during review   

It would be easier to talk about this feature if this were corrected

[littledan]

@rwaldron @bterlson What do you think?

[caitp]

per discussion with Dan:

• V8 currently ships the ES6-specified early ReferenceError behaviour (so, ReferenceError for --!x even in dead code) --- presumably this doesn't break the web or there would be people upset about this.
• At some point, this proposal said prefix-count operators had an LeftHandSideExpression operand rather than a UnaryExpression operand. If this were the case, --!x would become a SyntaxError rather than a ReferenceError.
• This seems to be changed back to its original (ES6) state --- but there is some inconsistent text between those two URLs mentioned above, which makes it unclear.
• Either way, this is an early error, and either way, it basically matches what already happens, so it shouldn't break the web

The only issue here is fixing the editorial bug, in order to clarify whether --!x becomes a SyntaxError, or remains an early ReferenceError.

I am fairly confident that the proper change is fixing http://rwaldron.github.io/exponentiation-operator/#PostfixExpression to match http://rwaldron.github.io/exponentiation-operator/#sec-update-expressions

Images of what I'm talking about, because it seems to be very hard to convey this:

http://rwaldron.github.io/exponentiation-operator/#sec-update-expressions:

http://rwaldron.github.io/exponentiation-operator/#PostfixExpression:

In other words, it's just a little editorial issue that makes it hard to talk about this feature during code review