Re: Task Queue Ambiguity

Hi Essam,

Essam K <essam.k@outlook.com>, 2015-05-18 20:23 +0000:
> Archived-At: <http://www.w3.org/mid/BAY181-W101040A0D714BC5344308BF0C40@phx.gbl>
> 
> Hi Mike,
> Thank you for the answer. Can you refer some implementation person who might know about it?

For the stuff you’ve been asking about, I think a good person to chat with
is probably Yoav Weiss. I think the best place to find him is probably on
#respimg on irc.w3c.org during CET hours. Otherwise, (re)asking on #whatwg
on irc.freenode.net when there’s activity there might get some responses.

  —Mike

> Regards,Essam> Date: Tue, 19 May 2015 02:25:02 +0900
> > From: mike@w3.org
> > To: essam.k@outlook.com
> > CC: public-html@w3.org
> > Subject: Re: Task Queue Ambiguity
> > 
> > Hi Essam,
> > 
> > Essam K <essam.k@outlook.com>, 2015-05-17 00:54 +0000:
> > > Archived-At: <http://www.w3.org/mid/BAY181-W778C97A77F0E867A680D5AF0C50@phx.gbl>
> > > 
> > > Hi,
> > > I have two questions that are not clear from the specs
> > > - As per the specs
> > > (http://www.w3.org/TR/html5/webappapis.html#event-loops) there is one
> > > event loop per unit of related similar origin browsing contexts. Are the
> > > task queues between these browsing contexts shared or every browsing
> > > context has it's own separate queues? 
> > 
> > I think that in implementations they are not shared. I think in practice
> > implementations have one event loop per render process—so an event loop
> > isn’t shared among multiple render processes, and a task queue feeds only
> > one event loop. (Otherwise it’s hard to imagine how things would work.)
> > 
> > > - Directly reachable browsing context with different origin will have
> > > different event loop. I am assuming that in that case every browsing
> > > context will have it's own task queue. Is that correct?
> > 
> > Yeah, I think that’s correct.
> > 
> > I say all this with the caveat that I’m not someone who’s actually
> > implemented any of it, so you probably want to either hear from somebody
> > who has implemented it or who has a less fuzzy idea than I do about the
> > actual implementations do things.
> > 
> >   —Mike

-- 
Michael[tm] Smith https://people.w3.org/mike