On Mon, Nov 10, 2008 at 9:10 PM, Garret Wilson <[EMAIL PROTECTED]>wrote:

>
> Michael,
>
> Actually, while I was waiting I did a test---I can successfully cancel
> an upload, but I get no ready state change after I manually abort.


Right... in looking a our implementation it won't fire an readystatechanged
event in the abort case, although it should depending on the state at the
time the abort method is called...
http://www.w3.org/TR/XMLHttpRequest/#abort

   1.

   Abort the send()
algorithm<http://www.w3.org/TR/XMLHttpRequest/#abort-send-algorithm>,
   set the response entity
body<http://www.w3.org/TR/XMLHttpRequest/#response-entity-body> to
   "null", the error flag <http://www.w3.org/TR/XMLHttpRequest/#error-flag> to
   "true" and remove any registered request headers.
   2.

   The user agent *should* cancel any network activity for which the object
   is responsible.
   3.

   If the state is UNSENT<http://www.w3.org/TR/XMLHttpRequest/#unsent-state>
   , OPENED <http://www.w3.org/TR/XMLHttpRequest/#opened-state> and the
   send() flag <http://www.w3.org/TR/XMLHttpRequest/#send-flag> is "false",
   or DONE <http://www.w3.org/TR/XMLHttpRequest/#done-state> go to the next
   step.

   Otherwise, switch the state to
DONE<http://www.w3.org/TR/XMLHttpRequest/#done-state>,
   set the send() flag <http://www.w3.org/TR/XMLHttpRequest/#send-flag> to
   "false" and synchronously dispatch a
readystatechange<http://www.w3.org/TR/XMLHttpRequest/#readystatechange>
event
   on the object.
   4.

   Switch the state to
UNSENT<http://www.w3.org/TR/XMLHttpRequest/#unsent-state>.
   (Do not dispatch the
readystatechange<http://www.w3.org/TR/XMLHttpRequest/#readystatechange>
    event.)




>
> Garret
>
> On Nov 10, 8:56 pm, Michael Nordman <[EMAIL PROTECTED]> wrote:
> > Hi Garret,
> > The request object transitions to the 'complete' ready state (4) and the
> > req.status property accessor will throw an exception after the request
> has
> > completed due to a network error or an explicit call to abort.
> >
> > On Mon, Nov 10, 2008 at 8:45 PM, Garret Wilson <[EMAIL PROTECTED]
> >wrote:
> >
> >
> >
> > > Austin,
> >
> > > Thanks for the reply, but it was terribly vague.
> >
> > > "...thus carries the same fate..."---what fate is that?
> >
> > > "...in a particularly direct fashion."---what particular fashion is
> > > it, then?
> >
> > > Perhaps the "fate" isn't "particularly direct", but surely *something*
> > > happens. I need to know what that "something" is, because it isn't
> > > documented.
> >
> > > * When a network error occurs, what does the readyState change to?
> >
> > > * When a network error occurs, is onprogress() called, and with what
> > > values?
> >
> > > * When I cancel an upload, what does the readyState change to?
> >
> > > * When I cancel an upload, is onprogress() called, and with what
> > > values?
> >
> > > Regardless of whether XHR is well documented, Gears is a Google
> > > produce that presumably has some deterministic behavior in the
> > > presence of errors. Why can't that be documented?
> >
> > > Garret
>

Reply via email to