Posted by Michael Fesser on 01/31/07 20:29

..oO(Erwin Moller)

>Michael Fesser wrote:
>>
>> 14.30 Location
>> http://www.w3.org/Protocols/rfc2616/rfc2616-sec14.html#sec14.30
>
>Hmm yes. That sounds convincing.
>Never trusts rfcs not recommended by me. ;-)
>
>Here is one for you:
>http://www.ietf.org/rfc/rfc3986.txt
>
>It clearly describes relative referencing. (Part 4 and 5 mainly)

I'm familiar with it, but even the most recent and generalized RFC for
URI syntax doesn't change the meaning of "absolute URI", which is
required in the HTTP RFC. RFC 3986 just defines what an absolute URI
actually is.

>> But you haven't seen all. And browsers are not the only user agents. A
>> bug doesn't become less of a bug, just because it works in the majority.
>
>Well, that may be true.
>I only aim my php for internetbrowsers, so I don't know jack about other
>agents.
>Thanks for the warning though.

I can remember someone having a strange problem with using a relative
URI and a query string in a Location header, which didn't work in Lynx
or something like that. The problem vanished after changing to an
absolute URI. It was posted to a German newsgroup, but I didn't bookmark
it and can't find it at Google Groups at the moment ... :(

Of course it doesn't really prove anything, but at least it shows that
there might be problems, which can easily be avoided by just following
the standards. Be it HTML, CSS or HTTP - it's always a good idea to
adhere to the written standards, because then it's not your fault if
something goes wrong.

Micha

[Back to original message]