[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [idn] Extending gethostbyname()




> Mark.Andrews@nominum.com writes:
> > Yes we are talking here about extending RFC 952, however
> > that does not imply that it is necessarily safe to extend
> > gethostbyname().  The safest thing to do is to specify a
> > new API which is defined to accept IHN and return IHN's.
> 
> Perhaps you have forgotten that RFC 952 is obsolete. RFC 1123 extended
> the host name syntax twelve years ago.

	RFC 1123 did not obsolete RFC 952, it updated it.

> 
> The gethostbyname() maintainers could have insisted that gethostbyname()
> be used only for the old syntax. They could have introduced an idiotic
> new gethostbyextendedname() function. They could have forced programs to
> switch from gethostbyname() to gethostbyextendedname() to support the
> new syntax.

	I'm sure they thought about it.  Relaxing the requirements
	to allow a digit as the first character of a label broke
	some applications as did the extention from 63 to 255
	characters.  These were trivial changes compared to what
	is being proposed here.

	It took years for some mailers to be able to mail to
	user@3com.com.

> 
> In short, they could have changed a simple syntax extension into a
> compatibility disaster. But they didn't.

	This is not a simple syntax extension.  It a radical change.

	Mark

> 
> ---Dan
> 
--
Mark Andrews, Nominum Inc.
1 Seymour St., Dundas Valley, NSW 2117, Australia
PHONE: +61 2 9871 4742                 INTERNET: Mark.Andrews@nominum.com