W3C home > Mailing lists > Public > public-webapps@w3.org > July to September 2012

RE: [XHR] Event processing during synchronous request

From: Mike Wilson <mikewse@hotmail.com>
Date: Wed, 12 Sep 2012 20:35:15 +0200
Message-ID: <SNT129-DS17A3C55B286FF5629AF7BAA4920@phx.gbl>
To: "'Anne van Kesteren'" <annevk@annevk.nl>
CC: <public-webapps@w3.org>
Anne van Kesteren wrote:
> On Mon, Sep 10, 2012 at 10:05 AM, Mike Wilson 
> <mikewse@hotmail.com> wrote:
> > Thanks. So I understand it is the HTML specification reference
> > that provides details on this behaviour for the XHR spec.
> > I'll look further there as its fetching section by itself
> > doesn't directly express the relationship between synchronous
> > and a stopped event loop (although as a programmer I can infer
> > this relationship).
> 
> Well yeah, shit's complicated. HTML fetch can be invoked either
> synchronously or asynchronously. When invoked synchronously it's as
> part of some task already and since the event loop is not manually
> spun no other tasks will run at that time. When invoked asynchronously
> HTML fetch itself queues tasks on the event loop that then cause
> various things to happen per XMLHttpRequest, such as dispatching
> progress events.

Thanks, and yes, for me this section:
http://www.whatwg.org/specs/web-apps/current-work/multipage/webappapis.html#
event-loops
has the sufficient text to define this.

On the other hand, in:
https://bugzilla.mozilla.org/show_bug.cgi?id=785243
Firefox implementors say that:
  "sync XHR spins the event loop, so events (including async 
  script execution and whatnot) can fire under a sync XHR call."
which seems to suggest that the XHR spec should allow 
implementations to pump events during a synchronous fetch.

What is the XHR spec teams's opinion on this?
Best regards
Mike
Received on Wednesday, 12 September 2012 18:35:51 GMT

This archive was generated by hypermail 2.3.1 : Tuesday, 26 March 2013 18:49:54 GMT