lynx-dev
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: lynx-dev Lynx, relative links, and the HTTP redirections


From: David Woolley
Subject: Re: lynx-dev Lynx, relative links, and the HTTP redirections
Date: Wed, 13 Jan 1999 23:39:00 +0000 (GMT)

> There is no problem with Netscape 4, which computes the relative links in 
> respect with the redirected document.

I think you are in a rather grey area.  The only way that Netscape can
fix up the relative URL is by looking at a header which indicates the
true URL returned, however, if the error is reported correctly, the
HTML is not the content of the resource, but an explanation of why the
resource is not available; one interpretation, if the header is present,
would be that the error page URL is the one that is missing!  If it's
done wrongly, as WILL happen if the error URL is non-local, automated
tools will get confused.

I'm not sure that the relevant header is defined in HTML 1.0, but could
be wrong.  If it isn't, the returned HTML can't safely rely on the browser
being aware of the true base.

> Which browser is right? How can I use relative links and the ErrorDocument 
> feature?

My advice would be to use server relative, not page relative URLs.
I think even better advice would be to steer clear of such error pages, as
they usually omit information that is in the builtin page.  Particularly
useful, if provided, are the failed URL and the referer, as these can
make it easy to email the response to the owner of a broken link.

Incidentally, some IE5 betas don't display the error page, but display
their own message when they get a 404 type return - probably to do
with the dumbing down of the user interface - such pages are probably
deemed too technical and it is easier to support a single, internally
generated message.

reply via email to

[Prev in Thread] Current Thread [Next in Thread]