Resolver Issues (non valid hostname characters)

[Mark.Andrews at isc.org]

> On Tuesday, March 25, 2003, at 08:46  PM, Terry Lambert wrote:
> 
> > Marius Strom wrote:
> >> I've submitted a PR for this: misc/50299 documenting the RFC
> >> mis-following (is that a word?) as well as a patch for res_comp.c.
> >
> > Great.
> >
> > If this is committed before RFC-952 is updated, FreeBSD users
> > can now define host names that break other machines on the net
> > which are strictly conformant to RFC-952.
> 
> Which will just make us behave like rest of the world.  I have tested
> resolvers on Solaris, Windows, MacOS X, MacOS 9, IRIX, Linux, AIX.
> They all will resolve a name with an underscore character.  Only the
> *BSD boxes fail because of the check.
> 
> > What is the first maxim of protocol design?
> >
> > "Be generous in what you accept, strict in what you generate".
> 
> Which is why I would argue that the patch should be committed, maybe
> with an option to enable it.  We are talking about the resolver, not a
> DNS or hostname server.  The resolver should resolve the name, be
> generous.  It just depends at what level you apply the maxim.  The
> check should be in the DNS server not in the resolver IMHO.
> 
> DaveD

	The current implementation fits this.  It handles (accepts)
	garbage in and only returns (generates) clean respones to
	the application.

	If the resolver died receiving underscore you would something
	to complain about.  Currently it just filters out ALL illegal
	responses.

	Mark
--
Mark Andrews, Internet Software Consortium
1 Seymour St., Dundas Valley, NSW 2117, Australia
PHONE: +61 2 9871 4742                 INTERNET: Mark.Andrews at isc.org