Hostname Naming Compliance
kcd at chrysler.com
Fri Feb 27 02:53:51 UTC 2009
Matus UHLAR - fantomas wrote:
>> Mark Andrews wrote:
>>> When does it stop? What will be the next character you
>>> "just have to have"? At the moment you have 1 inter label
>>> seperator and 1 intra label seperator. That should be
>>> enough for anyone.
> On 25.02.09 08:49, Peter Laws wrote:
>> Like 640k of memory.
> the main effect of allowing underscores would be that some companies would
> want/need to buy much more domains, e.g.
> I don't see any benefit in that.
>> Unicode is coming (as fast as IPv6, maybe faster :), so maybe it /is/ time
>> to update the naming standards.
> and maybe it is not. If people can't behave, adjusting standards may be the
> worst solution.
But, as far as I can tell, there's no *practical* reason to disallow
underscores, other than the fact that it may trip the standards-checking
code of some _other_ piece of software. So, piece of software A
disallows underscores because it's worried about causing a problem for
piece of software B, and piece of software B keeps the restriction
because it's worried about about causing a problem for piece of software
C, and piece of software C keeps the restriction because it's worried
about causing a problem for piece of software A.
Do you see how self-defeating that is? Everyone is looking out for
everyone else, yet there is no actual *real* problem with allowing
underscores. They're all just trying to protect each other against an
I've heard that in the old old days (70s, perhaps earlier) some
teletypes had a problem distinguishing between an underscore and a
backspace. That was a real honest-to-goodness *problem* with
underscores, and is probably why underscore was banned from hostnames in
the first place. But those teletypes are long gone. Rusted away or in a
museum somewhere. Get over it.
I agree with not changing standards to accommodate "bad behavior". But,
at the same time, the standards need to have a practical basis, not be
arbitrary or just a carryover from decades ago. As far as I can tell,
the underscore restriction, in particular, is just a legacy carryover
and has no practical use.
More information about the bind-users