On 03/17/99 22:07:19 you wrote: OTOH, are non-doted IP adresses valid URLs?
They are not required according to the RFC (see guy's email for a full discussion). When I did a quick look I couldn't find any other interface documents (telnet, etc.) that required anything other than dotted decimals for IP translation, either, but I can't say for certain that's entirely true. Anyway, when it comes to URLs, the fact that they work is an artifact of the code, not the specification.
Bruce
At 12:28 am -0600 18/3/99, sirbruce@ix.netcom.com wrote:
On 03/17/99 22:07:19 you wrote: OTOH, are non-doted IP adresses valid URLs?
They are not required according to the RFC (see guy's email for a full discussion). When I did a quick look I couldn't find any other interface documents (telnet, etc.) that required anything other than dotted decimals for IP translation, either, but I can't say for certain that's entirely true. Anyway, when it comes to URLs, the fact that they work is an artifact of the code, not the specification.
I'm not so sure they're a bad artifact though. If you start from the premise that you should be allowed to enter an IP address as a URL, and consider that in fact an IP address is defined as an unsigned 32-bit integer, then it sort of makes sense to allow it. After all, dotted quad notation was simply an easier way to write the 32-bit integer...
Having said that, does anyone have the latest URL specification to hand? If that says "dotted decimal", I'll shut up, but if it just says "IP address" I reckon that lack of support could be classified as suboptimal.
Alex
"During my service in the United States Congress, I took the initiative in creating the Internet." -- Al Gore, 9 March 1999
I'm not so sure they're a bad artifact though. If you start from the premise that you should be allowed to enter an IP address as a URL, and consider that in fact an IP address is defined as an unsigned 32-bit integer, then it sort of makes sense to allow it. After all, dotted quad notation was simply an easier way to write the 32-bit integer...
RFC 1738 says in section 3.1:
host The fully qualified domain name of a network host, or its IP address as a set of four decimal digit groups separated by ".".
I think you -- or rather, your users -- may be out of luck. :-(
Nick Hilliard Ireland On-Line System Operations
At 11:23 am +0000 18/3/99, Nick Hilliard wrote:
I'm not so sure they're a bad artifact though. If you start from the premise that you should be allowed to enter an IP address as a URL, and consider that in fact an IP address is defined as an unsigned 32-bit integer, then it sort of makes sense to allow it. After all, dotted quad notation was simply an easier way to write the 32-bit integer...
RFC 1738 says in section 3.1:
host The fully qualified domain name of a network host, or its IP address as a set of four decimal digit groups separated by ".".
I think you -- or rather, your users -- may be out of luck. :-(
Fair enough! At least now I have something to say to them....
Alex
"During my service in the United States Congress, I took the initiative in creating the Internet." -- Al Gore, 9 March 1999
On Thu, 18 Mar 1999, Alex French wrote:
I'm not so sure they're a bad artifact though. If you start from the premise that you should be allowed to enter an IP address as a URL, and consider that in fact an IP address is defined as an unsigned 32-bit integer, then it sort of makes sense to allow it. After all, dotted quad notation was simply an easier way to write the 32-bit integer...
OTOH, why should it support the non-doted decimal representation of the IP address. I'd like to see support for hex, octal, and binary at minimum. Decimal is not as important to computers.
Tom
-
Alex French -
Nick Hilliard -
sirbruce＠ix.netcom.com -
tkaczma＠gryf.net