|
|
Posted by Alan J. Flavell on 07/31/06 11:44
On Mon, 31 Jul 2006, Els wrote:
> Alan J. Flavell wrote:
>
> > Do I have to repeat that the server does not merely "add the
> > slash", but has to send back a redirection transaction (which
> > would otherwise have been unnecessary)?
>
> I was only describing what I saw as a user, to distinguish between
> the two cases.
That does, however, depend on the browser. E.g I've got a version of
Opera here on my office computer (the latest version is installed at
home), and if I go to Tools> Preferences> Network, I can un-check the
"Enable automatic redirection" option (which is enabled by default,
and then, when I need to be redirected, I get the "short hypertext
note" as it calls it in RFC2616, e.g at
http://rfc.net/rfc2616.html#s10.3.2 (second paragraph), thus alerting
me to the diversion.
> Sorry, I should have used the words "a slash is added"
> instead of "the browser will add the slash".
I think the key point is that the *server* will initiate a redirection
transaction to the corrected URL, and that this in general cannot be
avoided, due to the need to resolve relative URLs correctly. Thus, an
unnecessary additional network overhead has been inserted when any
author decided to rely on this server correction.
I think any answer which doesn't say that directly, is going to whoosh
over the head of anyone who hasn't already met this poblem area.
Navigation:
[Reply to this message]
|