http://3505021947 is an invalid URL

[Lawrence Chan]

Glen Turner wrote:
> 
> The definition of a URL only allows the dotted-quad form
> of the IP address.
> 
> So it is perfectly valid for a browser or HTTP proxy to
> reject the URL as invalid.
> 
> RFC1738 section 5 says:
> 
>   host           = hostname | hostnumber
>   hostnumber     = digits "." digits "." digits "." digits
> 
>                -------------------------------
> 
> Most implementations hand <host> down to gethostbyname(),
> if that fails they then try inet_addr().  inet_addr() allows
> a lot of formats for the IP address.  This is actually
> a really bad idea.  For example the strings:
> 
>   010.010.010.010
> 
> and
> 
>   10.10.10.10
> 
> are different addresses.  Some UNIXen are more restrictive
> about the formats accepted by inet_addr() to solve the user
> interface problems of the more generalised BSD implementation.
> Compare DEC UNIX to Solaris.
> 
> Cheers,
> glen
> 
> --
>  Glen Turner                               Network Specialist
>  Tel: (08) 8303 3936          Information Technology Services
>  Fax: (08) 8303 4400         The University of Adelaide  5005
>  Email: glen.turner at adelaide.edu.au           South Australia

Hello,

Somebody just showed me another interesting URL format which starts with
a set of %dd, the d seems dec.  The URL comprises of 9 of these sets at
the beginning followed by an @ and another 10 digits, also appear to be
dec like:

%dd%dd%dd%dd%dd%dd%dd%dd%dd at dddddddddd

What would you think of that?  How does one find out who the URL belongs
to?

Lawrence Chan
lchan at montevino.com