Re: Redirect-Request

To: CHRISTOPHER ELIOT <eliot@cs.umass.edu>
Subject: Re: Redirect-Request
From: "John C. Mallery" <jcma@ai.mit.edu>
Date: Mon, 24 Feb 1997 17:57:23 -0500
Cc: www-cl@ai.mit.edu
In-Reply-To: <Pine.ULT.3.94.970224163802.25053B-100000@themis.cs.umass.edu>
Sender: owner-www-cl@ai.mit.edu

At 4:56 PM -0500 1997-02-24, CHRISTOPHER ELIOT wrote:
>Folks,
>
>Redirect-request and :redirect URLs don't work when *debug-server* is T
>because they signal a condition which is then not handled and becomes an
>error.
>
>I find this inconvienient. Is it considered a bug or a feature?

People come down on both sides of that one. Try the mac devo 60.104 and
see if you like that any better. I believe I left in bind *break-on-signals*
so that you can proceed from conditions like redirects. People moaned about
this approach on other platforms like the lisp machine, which signals tons
of stuff.

>
>I'm trying to make a URL use an outside server when I'm connected to the
>internet, while keeping a summary of that material available when I do
>demos in standalone mode on my powerbook. I'm talking about a link to my
>research group's home page. The link is defined in an image map, so I must
>create a single URL that works in both cases.

Best to just copy the example of a computed redirect in: 

http://wilson.ai.mit.edu/cl-http/sources/examples/exports.lisp

Change the criteria from browser capabilities to whether you're connected.

References:

Redirect-Request

From: CHRISTOPHER ELIOT <eliot@cs.umass.edu>