W3C home > Mailing lists > Public > whatwg@whatwg.org > April 2012

[whatwg] readyState transitions when aborting a document

From: Henri Sivonen <hsivonen@iki.fi>
Date: Thu, 19 Apr 2012 14:43:12 +0300
Message-ID: <CAJQvAue7-RRkXTGf+cOZ6raBwDi1dJCtwj5js_qfyKBfjvjqkg@mail.gmail.com>
I've been trying to make document.readyState transitions less broken
in Gecko. (The transitions are very sad as of Firefox 13 in pretty
much all but the most trivial cases.)

I'm having a particularly hard time figuring out what the right thing
to do is when it comes to aborting document loads. Unfortunately, I
don't trust the spec to describe the Web-compatible truth.
 * Is there a way to abort a document load in IE without causing
immediate navigation away from the document? IE doesn't support
window.stop().
 * Does Web compatibility ever require a transition from "loading" to
"complete" without an intermediate "interactive" state?  (Both chrome
and Firefox as shipped make such transitions, but those might be
bugs.)
 * Should the aborted documents stay in the "loading" state forever
like the spec says or should they reach the "complete" state
eventually when the event loop spins?
 * Should window.stop() really not abort the parser like the spec
seems to suggest?
 * Should reaching "complete" always involve firing "load"?
 * Should reaching "interactive" always involve firing "DOMContentLoaded"?
 * Does anyone have test cases for this stuff?

-- 
Henri Sivonen
hsivonen at iki.fi
http://hsivonen.iki.fi/
Received on Thursday, 19 April 2012 04:43:12 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 30 January 2013 18:48:07 GMT