Number.NaN and reserved word

[kkapsner]

If I use Number.NaN (which is perfectly valid) I get

Expected an identifier and instead saw 'NaN' (a reserved word).

First: NaN is NOT a reserved word - it is a global Variable that should not be overwritten (and JSHint should warn you if you do it).
Second: Number.NaN is part of the ES-spec (in 3 and 5)

PS: in ES3 are some FutureReservedWords (like "boolean") which are not recognised in JSHint. If I write

var obj = {boolean: true};

I get no warning form JSHint although it's invalid ES3 (not ES5), because in ES3 an identifier cannot be a reserved word.
In fact I know no Browser who claims about this (I only checked "boolean") - but just for consistency.

[WolfgangKluge]

Yes, it's a bug. But can you give me an example where you use Number.NaN.
As I understand, you can only use it for assignments, not for tests (that's what isNan is for). Right?

var x = Number.NaN;
if (isNaN(x)) {
    x = 2;
}

Reserved keywords.
The future reserved keywords in ES3 are

abstract enum int short
boolean export interface static
byte extends long super
char final native synchronized
class float package throws
const goto private transient
debugger implements protected volatile
double import public

In ES5 the list ist somewhat shorter

class enum extends super 
const export import

when in strict-mode, the following are future-reserved, too:

implements let private public yield
interface package protected static

I think, jshint should warn if I use one of them, too - even if they are defined in ES3 or if the parser is not in strict-mode.

[kkapsner]

Yes, I use Number.NaN for an assignment (if any numeric property can not be destiguished I assign Number.NaN - I could also just use NaN... BUT in older Browser it may be overwritten by a different script) and you can not use it for testing because

(NaN === NaN) === false

only value in JS that is not equal to it self.

[kkapsner]

The "boolean"-issue is still open...