[OpenID] "Free" URLs and XRIs (was RE: why is xri so obtuse?)

[Bob Wyman]

On 1/2/07, Drummond Reed <drummond.reed at cordance.net> wrote:
>
> Right – it's never quite clear what folks mean when they say "free" in
> this context. For example, a common misconception is, "URLs are free but
> XRIs cost money."
>

I think an important point is being missed here... It would be useful, I
think, to consider the fact that global XRI's represent an *incremental*
expense that has not been traditionally required. Arguments that merely
point out that other things, such as domains, cost money or that emphasize
similarity in pricing structures do not effectively address the issue of
*incremental* expense.

Traditionally, Internet users have only been aware of three things that need
to be paid for (by someone) in order to get full use of the Internet: 1)
Domain names, 2) Bandwidth, 3) Client computers.  (Remote machines that
implement services are treated as though they are "free" since someone else
pays for them.) Once the three resources have been obtained, an Internet
user has traditionally been able to use most of the Internet's resources
without additional incremental expense. Importantly, users have also been
able to create new "names" freely since those "names" have traditionally
been rooted by the domain names which have *already* been paid for.

Now, XRI comes and suggests that names need to be paid for... Even though
the charges may be low, even reasonable, it is important to understand that
if only three things have been paid for in the past, then adding a fourth
thing is a major event. We shouldn't be surprised to discover some push-back
when the number of items that need to be paid for increases by 33%. (4/3=
1.33) We also shouldn't be surprised if people are wary of the various
suggestions that "free" names can still be had (based =freeid, etc.) since
the user is still suddenly aware that someone had to pay for the =freeid
base name... The shape of the universe has changed.

Given a focus on incremental expense rather than relative or fair pricing, I
think its fairly obvious that the issue is not as simple as Drummond Reed
presents it. The issue isn't that "URLs are free but XRIs cost money." The
issue is that traditionally, "Domains cost money but all other names are
free." Since an XRI is a name which is not a domain, it seems obvious to
many that it should be free.

This issue of incremental cost is, I think, particularly grating when it is
raised in the context of OpenID. The reason is that OpenID, for many people,
was expected to follow the path of relying on "free" names which are
admittedly based on those very special names (domains) that everyone knows
you must pay for. It was expected that OpenID would provide names that did
NOT require incremental expense. Some of us, for instance, have been of the
opinion that one of the greatest hindrances to the adoption of effective
identity technology has been the fact that most technically viable systems
have become practically non-viable simply because they require that people
assume incremental and often recurring expenses in order to obtain identity
means. (e.g. Certificates purchased from Verisign, Thawte, etc...) Now,
OpenID+XRI comes and presents us with the specter of incremental expense for
identity means... One should not be surprised by the push-back. As anyone in
marketing will tell you: Whether or not the push-back is "reasonable" is
not, unfortunately terribly relevant. ("Reality is rarely relevant, it is
perceptions that matter...")

My intent in this note has not been to attack XRI. Rather, I'm just trying
to illuminate what I believe to be the source of at least some of the
resistance. The problem here is that arguments on both sides of the issue
appear to be quite rational and even reasonable. When the XRI folk say,
"It's just like domains." They are right. But, when the opposition objects
to the incremental expense, they are also right.

My personal opinion is that this controversy is not good for OpenID. No
matter how wonderful and excellent the XRI scheme may be, its incorporation
into OpenID seems to be producing resistance, confusion and complexity that
will make it more difficult to get OpenID accepted. The risk is that we will
be left where we are today -- with many great ideas for identity technology
but no effective means to get the stuff adopted.

bob wyman
(reachable as =bobwyman <http://xri.net/=bobwyman> as long as I keep paying
for the privilege...)
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openid.net/pipermail/openid-general/attachments/20070102/1e767be7/attachment-0002.htm>