Re: Re: Does Location: headers constantly

>> I can't let this go unchallenged -- what happens if you're sending this to a >> browser which *doesn't* honour the "Location: " header, but *is* capable of >> displaying any attached page content? The poor user will be left with a >> completely blank screen, and no hint of why. In fact, the HTTP/1.1 spec > >But, which browser is incapable of handling location headers? I think Netscape 1.x didn't... I'm pretty sure Netscape 2.x did support Location headers. At least, that's my recollection of their behaviour. Now, find me an actual user who is really *using* Netscape 1.x... :-) There might be some really stupid search engines that don't, but no human will see the text anyway... I suppose a *really* stupid search engine could ignore the headers and parse the HTML... :-) By all means, if you have valid HTML that you want to send just in case they are using Fred's Browser that doesn't handle Location headers, do so. I only meant that you shouldn't have stupid code that sends the rest of your HTML that's just sort of sitting there after your Location:, and that Cookies are just not going to work: ----------- bad code ------------------------------------ <?php if (!valid_user()){ header("Location: http://www.example.com/login.htm"); } ?> Secret information here. --------------------------------------------------------- I don't know really care what the RFC's say -- I know what *WORKS* in the browsers. Alas, the browser developers are racing each other to release, and they ain't reading the RFC's either. :-( Maybe the RFC's say that Cookies should work. They don't work. Sorry. When the W3C starts promulgating standards that are actually *IMPLEMENTED* *RELIABLY*, I'll start paying more attention to them. Meanwhile, I'm coding for the actual behaviour of the browsers in use: Netscape/IE version 3 through 6 (7?), Opera, Mozilla, and Lynx. No CSS. No layers/DHTML. Minimal JavaScript for roll-overs. No Flash. I dropped support for version 2 browsers only last year or so when a client dumped her 486 (386?) and got a new laptop... That was a relief. If you want challenges in supporting obsolete systems, try building web-sites for starving artists. They use the hardware their day-jobs throw away :-) >> (RFC 2616: >> http://www.w3.org/Protocols/rfc2616/rfc2616.html) even says: >> >> > Unless the request method was HEAD, the entity of the response >> > SHOULD contain a short hypertext note with a hyperlink to the >> > new URI(s). > >Oddly, I don't find that text you excerpted on the URI you provided. > >But, the thing is, I'm under the impression that PHP (and/or Apache?) >doesn't send anything else from your script once a Location header is >sent. I just confirmed this with two different tests. Maybe RFC 2616 got changed?... I dunno. Or care. For browsers, I have to work with reality, not RFC's. I sure *wish* they matched up, like they do for most other things (or pretty close, anyway) but they don't. -- Like Music? http://l-i-e.com/artists.htm