Re: XRI vote aftermath

Hash: SHA1

Schleiff, Marty writes:

> I'll start by asking for help to understand the TAG's stance on
> introduction of new URI schemes. I understand the part about it being
> costly to introduce new schemes. What I wonder about is the idea that
> the http: scheme should be used for everything (I probably didn't state
> that very well - perhaps you can put it into better words).
> If there existed no mailto:, or ldap:, or https: scheme today (the three
> I'm most familiar with beyond http:), what would be the TAG's reaction
> to a request for a new scheme for mailto: or ldap: or https?

I am not speaking for the TAG, only myself, in this, but the core of
my answer is in fact from the IETF [1]:

  "[T]he unbounded registration of new schemes is harmful.  New URI
   schemes SHOULD have clear utility to the broad Internet community,
   beyond that available with already registered URI schemes."

I simply don't think that's true for XRIs.  In trying to explain this
to someone else, before reading your email, I actually used Boeing as
an example:  I believe it's the case that most desktops in Boeing (and
there are a _lot_ of them) are centrally managed and tightly
constrained, with a multi-year roll-out cycle.  That means that no-one
at Boeing will be able to click on an xri: or hdl: or doi: URI for at
_least_ three years, given that IE7 does not support any of those out
of the box.  "So," I said to my interlocutor, "do you really want to
recommend that your users use URIs which no-one at Boeing, or dozens
of other similar companies, can click on for years to come?"

Yes, this is an argument against _any_ new URI scheme where there is
real value to be gained by allowing as many people as possible to use
it to access resources on the Web.  And the network effect (because
that's what we're talking about) is what _made_ the Web.  Using a new
URI scheme when you could use http: is intentionally cutting yourself
off from the network effect.

I think that mailto: is pretty clearly _not_ in that category, and
its non-retrieval semantics makes it a reasonable special case.  And
https: is really just http: with a bit of metadata encoded in the
scheme name.  ldap: is a less clear case -- I'm not really familiar
with its operation, but my superficial understanding is that it is not
central to the functioning of the LDAP system, and that its semantics
mean that it is unlikely that it will appear in general-purpose Web
contexts, which distinguishes it pretty well from http:.

The TAG is working on a detailed exposition of its position in this
matter, which I expect will address your question in more detail -- I
hope this quicker personal reply will be helpful in the meantime.


- -- 
 Henry S. Thompson, HCRC Language Technology Group, University of Edinburgh
                     Half-time member of W3C Team
    2 Buccleuch Place, Edinburgh EH8 9LW, SCOTLAND -- (44) 131 650-4440
            Fax: (44) 131 650-4587, e-mail:
[mail really from me _always_ has this .sig -- mail without it is forged spam]
Version: GnuPG v1.2.6 (GNU/Linux)


Received on Thursday, 5 June 2008 21:54:13 UTC