- From: Yves Lafon <ylafon@w3.org>
- Date: Tue, 25 Jan 2000 18:57:25 +0100 (MET)
- To: Tu Yu-Guang <tuyuguan@comp.nus.edu.sg>
- cc: www-jigsaw@w3.org
On Tue, 25 Jan 2000, Tu Yu-Guang wrote:
> Thank you very much. I would like to confirm with you so that I can ensure
> that I understand you correctly. Do you mean the client browsers such as
> netscape or ie are taking care of detecting infinite redirection loops, so
> nothing needs to be done on the web server side? also the maximum of five
> redirections are implemented by the browsers? all latest version browsers
> support this? If the client are not able to detect it, what do you suggest
> to do?
As HTTP is stateless, you won't be able to break loops server-side (unless
you add some cookies, but you can't rely on broken clients to use them),
so the best is to avoid doing loop redirection server side, if you can.
If the client is not able to detect the loop, it is difficult to handle
this in a clean way, forbiding temporarily an IP doing too many request
may be an option, but it may interact badly with "good" clients doing so.
Regards,
/\ - Yves Lafon - World Wide Web Consortium -
/\ / \ Architecture Domain - Jigsaw Activity Leader
/ \ \/\
/ \ / \ http://www.w3.org/People/Lafon - ylafon@w3.org
Received on Tuesday, 25 January 2000 12:55:49 UTC