Peter Mogensen <[email protected]> writes:
> On 2012-09-08 08:23, Russ Allbery wrote:

>> Sure, but you are.  When you do a GET to a WebAuth-protected site with
>> an expired token, you get a redirect to WebLogin followed by a redirect
>> back to the original site, which re-establishes your cookies and then
>> gives you the resource you asked for.  As long as your AJAX follows
>> those redirects, and for a GET there's no reason why it wouldn't,

> But following those redirects make you loose JavaScript state - which
> doesn't make web developers happy.

Oh, right, because it puts stuff in anchors that would be stripped by a
redirect.  Right, right.

> Problems such as this really makes me think HTTP authentication
> standardization have failed.

Yes, indeed.  All of this stuff is effectively just workarounds for the
lack of real HTTP authentication.  It's pretty frustrating, isn't it?

-- 
Russ Allbery <[email protected]>
Technical Lead, ITS Infrastructure Delivery Group, Stanford University

Reply via email to