W3C home > Mailing lists > Public > ietf-http-wg-old@w3.org > September to December 1995

Re: Decision about Host?

From: Rich Salz <rsalz@osf.org>
Date: Thu, 5 Oct 1995 13:27:45 -0400
Message-Id: <9510051727.AA16360@sulphur.osf.org>
To: efrank@ncsa.uiuc.edu
Cc: http WG <http-wg%cuckoo.hpl.hp.com@hplb.hpl.hp.com>
> 2) There is not portable way to easily guarantee that a
> 	client (on all platforms) can get a fully qualified
> 	domain name (fqdn) to place in HOST.

You can determine the FQDN of the server you just connected to.  I appreciate
that this can be awkward, if you have to slip a getpeername() call into
the lower-level library and later on do gethostbyaddr.)

You can determine if what you have is a FQDN (turn of resolver "searching",
do the lookup, turn it back on.  I don't know if you want to rely on BIND
existing and I don't know what other knobs other DNS clients give you).
If not, then proceed:  strip off as many sub-domains as have in the "bad"
name.  (E.g., if the FQDN of the last server is foo.ncsa.uiuc.edu and you
have bar.cs, then remove foo.ncsa).  Prepend the bad stuff (e.g., giving
you bar.cs.uiuc.edu) and treat that as the FQDN.

If this doesn't work, then the host's DNS stuff is so screwed up that it
is a wonder they can communiate at all. :)
	/r$
Received on Thursday, 5 October 1995 11:02:31 EDT

This archive was generated by hypermail pre-2.1.9 : Wednesday, 24 September 2003 06:31:33 EDT