XMLHttpRequest in IE7 with REPORT method

View: New views
4 Messages — Rating Filter:   Alert me  

XMLHttpRequest in IE7 with REPORT method

by Cyrus Daboo-2 :: Rate this Message:

Reply to Author | View Threaded | Show Only this Message

Hi,
Has anyone tried using XMLHttpRequest in IE7 to do WebDAV operations?
Specifically we are finding that the REPORT method seems to fail on the
.open() call (same thing for CalDAV's MKCALENDAR). PROPFIND works fine. Any
ideas on whether REPORT can be made to work?

We have also tried using the Msxml2.XMLHTTP ActiveX object. Whilst that
seems to accept REPORT as a valid method, it fails with digest
authentication (basic works). What we see is that it makes an
unauthenticated requests and gets back a 401 with the digest challenge, but
then it makes another unauthenticated request and fails (instead of sending
the digest response). Basic authentication works as does digest with
PROPFIND. Any ideas what might be up?

Is there any other way in IE7 to make WebDAV REPORT requests that will work?

--
Cyrus Daboo



Re: XMLHttpRequest in IE7 with REPORT method

by Julian Reschke :: Rate this Message:

Reply to Author | View Threaded | Show Only this Message

Cyrus Daboo wrote:
> Hi,
> Has anyone tried using XMLHttpRequest in IE7 to do WebDAV operations?
> Specifically we are finding that the REPORT method seems to fail on the
> .open() call (same thing for CalDAV's MKCALENDAR). PROPFIND works fine.
> Any ideas on whether REPORT can be made to work?

Only by falling back to the ActiveX variant of XmlHttpRequest.

This issue has been reported to Microsoft something long ago (see thread
around
<http://lists.w3.org/Archives/Public/public-webapps/2009JanMar/0352.html>),
but first they claimed it was "by design", and then it was too late to
change, even for IE8.

BTW: don't try to do this with Opera, it will silently change the method
name to GET.

> We have also tried using the Msxml2.XMLHTTP ActiveX object. Whilst that
> seems to accept REPORT as a valid method, it fails with digest
> authentication (basic works). What we see is that it makes an
> unauthenticated requests and gets back a 401 with the digest challenge,
> but then it makes another unauthenticated request and fails (instead of
> sending the digest response). Basic authentication works as does digest
> with PROPFIND. Any ideas what might be up?
>
> Is there any other way in IE7 to make WebDAV REPORT requests that will
> work?

I've used the ActiveX object a lot in the past, but probably not with
Digest.

BR, Julian




Re: XMLHttpRequest in IE7 with REPORT method

by Cyrus Daboo-2 :: Rate this Message:

Reply to Author | View Threaded | Show Only this Message

Hi Julian,

--On May 14, 2009 7:15:17 PM +0200 Julian Reschke <julian.reschke@...>
wrote:

>> Has anyone tried using XMLHttpRequest in IE7 to do WebDAV operations?
>> Specifically we are finding that the REPORT method seems to fail on the
>> .open() call (same thing for CalDAV's MKCALENDAR). PROPFIND works fine.
>> Any ideas on whether REPORT can be made to work?
>
> Only by falling back to the ActiveX variant of XmlHttpRequest.
>
> This issue has been reported to Microsoft something long ago (see thread
> around
> <http://lists.w3.org/Archives/Public/public-webapps/2009JanMar/0352.html>
> ), but first they claimed it was "by design", and then it was too late to
> change, even for IE8.
>
> BTW: don't try to do this with Opera, it will silently change the method
> name to GET.

Has any effort gone into defining a standard for tunneling unsupported HTTP
methods over, say, POST? That's what we will end up doing if there is no
solution on the browser side - but then our webapp won't work with other
servers unless they support POST-tunneling too.

--
Cyrus Daboo



Re: XMLHttpRequest in IE7 with REPORT method

by Helge Hess :: Rate this Message:

Reply to Author | View Threaded | Show Only this Message

On 14.05.2009, at 19:28, Cyrus Daboo wrote:
> Has any effort gone into defining a standard for tunneling  
> unsupported HTTP methods over, say, POST? That's what we will end up  
> doing if there is no solution on the browser side - but then our  
> webapp won't work with other servers unless they support POST-
> tunneling too.

Its not a real standard, but maybe this would be an reasonable  
fallback option:

   X-HTTP-Method-Override

   http://code.google.com/apis/gdata/docs/2.0/basics.html

We do support that in our frameworks.

Greets,
   Helge
--
Helge Hess
http://helgehess.eu/