Message for Bind-users

Hans Hohenner hansh at valicert.com
Wed Jun 14 00:26:50 UTC 2000


Since this is clearly an important issue to you, my suggestion is to author 
a RFP proposing the change.  It's better to work within the system (such 
that it is, the Internet can be a pretty chaotic place) than go off on a 
tangent.   Better yet, why not write one to allow the  Unicode character 
set to be used... and include the underscore in this list.  I'm sure you'll 
please lots of people not in USA that way too, and maybe get enough of a 
consensus to push it through.

If that gets accepted by the IETF, then the bind authors will (most likely, 
I'm not one of those esteemed folks, and can't really speak for them...) 
incorporate the changes into the code base.  I'm quite willing to bet that 
their goal is to follow the standards that are in place, not arbitrarily 
change them.

Hans

At 08:02 PM 6/13/2000 -0400, Kevin Darcy wrote:
>flaps at dgp.toronto.edu wrote:
>
> > Kevin Darcy <kcd at daimlerchrysler.com> writes:
> > >McNair, Dan wrote:
> > >> Correct me if I am wrong, but the underscore is not a legal character
> > >> in any domain name.  This is not a limitation of BIND, it is a 
> restriction
> > >> imposed by the domain name standard.  My guess is that there are both
> > >> practical and historical reasons for the restriction.
> > >
> > >There is no "practical" reason other than "this is the standard we agreed
> > >to way back when and we're afraid to change it because then we might break
> > >some lazy programmers' code (possibly causing security holes, cancer,
> > >famine, or maybe even global thermonuclear devastation)".
> >
> > Is there some particular reason you think underscores *should* be permitted
> > in hostnames?
>
>It enlarges the available namespace and offers a higher degree of naming
>flexibility.
>
> > The standard way to produce something looking like a space in a hostname
> > is to use a hyphen.
>
>It's a moot point what is "standard" and what isn't; I thought we were
>discussing whether the standard actually makes any sense.
>
> > A working system doesn't need multiple ways to do the same thing.
> > Hyphens suffice.
>
>Well, we don't *need* DNS at all: everyone could just use dot-notation IP
>addresses, or, for that matter, strings of 1's and 0's. DNS, and underscores,
>enhance the human/computer interface, or at least are perceived to by many of
>the humans who use the interface. It's not a matter of what "suffices", but of
>whether the cost of the restriction, or its relaxation, outweighs the benefit.
>
>
>- Kevin
>
> >




More information about the bind-users mailing list