L
Lasse Reichstein Nielsen
dhtml said:
According to the citation from the example in the ECMA 262 specification:
"... for example, in the HTML document object model the window property
of the global object is the global object itself."
there appears to be something called the HTML document object model which
has property of the global object called "window".
What might that "HTML document object model" be?
It's *not* the W3C HTML DOM Level 1 or W3C DOM Level 2 HTML specifications,
as these do not specifiy a "window" property.
And the Gecko DOM specification says:
"DOM Level 0. Not part of any standard."
(http://developer.mozilla.org/en/docs/DOM:window.window)
I.e., there is no standard requireing the "window" property, and that
includes ECMA 262.
Or to put it differently: If a browser implemented ECMA 262 and, say,
DOM 2, but didn't have a global "window" property, would it be ECMA
262 compliant? If yes, then obviously the citation is *not* normative.
And it is "yes"!
In the environment where you tested it.
I'd expect all web browsers to satsify this for a long, long time,
for of backwards compatability, but it's not required by the ECMA
262 specification. It's a de-facto standard, not de-jura.
It's the expected behavior because that's the behavior that many, many
web-pages out there expects, so no browser creator would be stupid
enough to fail to comply. But it's not mandated by ECMA 262.
He never claimed they were in the setting where you were testing.
Only that you couldn't be sure they would be in all settings, or
at least not just based on the ECMA 262 text.
It is possible, but unlikely to happen in practice, to have a browser
that is ECMA 262 compliant and that does not have a global window
property.
/L