dns media type registration tree

[Keith Moore]

> > I'm not sure why people assume that just because they are familiar
> > with how to allocate names in a particular name space, that that
> > name space should then be adapted to every purpose that comes to
> > mind.  There is value in having semantics associated with a name
> > space - a value which is diluted by using that name space for a wide
> > variety of purposes.  DNS is too overloaded as it is.
> 
> Agreed.  The dns tree doesn't remove the need for the IETF tree, nor
> the need for review.  IMO, it just does two very useful things; 
> 
> - it's a better form of x/vnd/prs tree, especially when combined with
> my proposed URI extension

How does making it easier to create new types equate to "better"?
 
> - it enables new authorities to establish their own review policies
> over their part of the media tree namespace

This is NOT inherently a good thing.  There aren't many restrictions in
the current IETF policy, and relaxing these few restrictions is probably
not in the best interests of the network.  Sure enough, some other
organizations  would also do due diligence in reviewing new types; some
would do a better job than IETF.  But encouraging anybody with a domain
name to register new types will certainly result in less review overall.
 
> > Beyond that, DNS is not well-suited for media types.  DNS
> > assignments are ephemeral.  They are subject to change as their
> > assignees (e.g. the organizations whose names they reflect) merge,
> > split, go bankrupt, fail to renew their registrations, or sell off
> > trademarks.  They are subject to reassignment for arbitrary reasons.
> > We discovered long ego that URIs based on DNS names are not suitable
> > for long-term (archival) use precisely because those names change;
> > that's why URNs and DOIs were invented.  And the utility of URNs has
> > been nearly destroyed by misuse and overloading of that name space.
> 
> Gotta disagree there.
> 
> There is *far* too much depending upon the DNS->authority mapping to
> have it ripped out from under us.  If something eventually replaces
> DNS(while still being centrally administered), there will be
> *enormous* and IMO, unescapable pressure to have that system be an
> extension of DNS so that existing DNS name->authority mappings
> persist.

You're talking about the DNS namespace as a whole; I'm talking about
individual DNS names.  I agree that there will be tremendous pressure
to maintain the DNS name space even if (say) the DNS protocol changes.
But we've seen numerous examples where DNS names were allowed to 
expire and were then reassigned; we've also seen a few examples where
DNS names were taken away from their original owners and reassigned for
apparently arbitrary reasons.
 
> I expect that I disagree with MarkN on this too; I don't think that
> the use of XML necessarily requires a multitude of media types (and
> I'm not talking about using "application/xml" for everything). 
> Personally, my money's on description frameworks like RDF to moderate
> the need for new media types.

Mumble.  Multi-layer dispatching seems like something to avoid; or at
least, to be wary of.

> So while I agree with you that a myriad of media types is bad, I don't
> believe that the DNS tree either encourages this (any more than do
> existing extensibility processes and mechanisms), nor do I believe
> that the tree is only useful in that case.

The existing mechanisms for defining new names try to strike a balance 
between the need for registration and the fact that people will define
their own types without registering them - if they don't understand how
to register new types or they believe it's too difficult to do so. 
This proposal would upset that balance and tip it in the wrong
direction.

> > If there really is a compelling need (meaning that it serves the 
> > greater good) to define new media types at a whim, a much better set
> > of names already exists, one which was more-or-less designed for
> > that purpose.  It's called OIDs.  They are easy to obtain.  They are
> > 
> > recursively extensible just like DNS names (actually moreso).  It is
> > 
> > relatively well-established that once assigned, the meanings of OIDs
> > do not change.  They don't contain human-readable content that
> > invites disputes over ownership.
> 
> They're also not derefenceable.  I think that's a huge loss.  See the
> draft mentioned above for more on that subject.

That's a feature, not a bug. 

> > The argument for DNS media types reminds me of countless other 
> > arguments for why protocol X should be used for everything (or at 
> > least, every instance of some large class of problem).   In the past
> > X has taken on values such as SOAP, XML, HTTP, SNMP, LDAP, URLs,
> > SSL, ASN.1, RPC, and even TELNET.  Most of those arguments look
> > pretty naive now, but people took them seriously when they were in
> > fashion.  Now it's essentially being argued that since DNS is a
> > widely-deployed namespace and query protocol, that it should be used
> > for yet one more thing that could be looked up.
> 
> AFAIK, it's the only universally deployed means of mapping strings to
> authorities on the Internet.  

In fact, that's not what DNS is.  It's a protocol for finding the
network addresses associated with hosts and services.

--
He not busy being born, is busy dying.		- Bob Dylan