[OpenID] W3C TAG recommends against XRI

[Gabe Wachob]

I tend to agree with Dick - the market should decide (or rather, it
*will* decide). In fact, I've stated this before in public.

That being said, lack of adoption of INames is not a reason to block
the vote in OASIS.

We can have a separate discussion of its role in OpenID - but I think
the focus we (the XRI TC) are having this week and next is the vote in
OASIS. The main objection has been that it "breaks" the web. We
vehemently disagree - in fact I think we've gone a long way to conform
to web architectural principles espoused by the W3C TAG.

I think one of the main problems with the XRI in OpenID discussion is
a confusion about new features in OpenID 2.0 somehow being the fault
of XRI. I think the new features (XRDS, directed identity, etc) would
have been implemented largely the same way. Many implementers are
ignoring XRI altogether (or simply relying on proxy resolution, which
I believe should actually be the model - there's not complete
agreement within the XRI community on that).

XRI causes a increased cognitive load for OpenID implementers for
sure. In terms of implementing OpenID RPs, I don't think there's
really additional code complexity by introduction of XRI, if people
are using proper libraries (which use proxy resolution).

I'm trying to turn the volume down a bit here, if you can't tell.
Reality isn't on either end of the spectrum, its in the middle.

    -Gabe


On Fri, May 23, 2008 at 1:38 PM, Dick Hardt <dick at sxip.com> wrote:
>
> On 23-May-08, at 4:09 AM, Drummond Reed wrote:
>
>> Dick, glad you asked. I was at an event all day today but spent a
>> good part
>> of this evening collating the input of XRI TC members in response to
>> the
>> TAG's posting.
>>
>> Suffice it to say, the XRI TC -- and so far the vast majority of OASIS
>> members -- believe the TAG is mistaken in its judgement.
>>
>> For the full details of our response, see:
>>
>>       http://wiki.oasis-open.org/xri/XriSolvesRealProblems
>
> Hmm.
>
> I think it would be useful to state why XRI is important without
> muddying the conversation with XRDS. For example in the document you
> list above in the Identity Capabilities Description, George Fletcher
> and Axel Nennker are talking about XRDS, not XRI.
>
> Clearly there is momentum around XRDS and it is resonating in the
> market. I am not seeing the same for XRIs.
>
> The recent IIW Presentation I gave on redundant identifier sets
> potentially solves the identifier recycling and loss of identifier
> problem in a manner that has advantages over XRI and does not require
> the creation of another central infrastructure. In other words, there
> are other ways to solve the problem without requiring the new
> infrastructure.
>
> Personally, I have not seen significant advantages to XRIs (and yes, I
> have a couple -- vanity at work!).
> I have taken the position that we should let the market decide if
> there is value in XRIs. If I look at the market, I think people in
> general see URLs being much simpler to work with then XRIs and.
>
> As a colleague and as someone that wants to see someone smart like
> yourself doing productive work, I'd suggest there may be value in
> reflecting on what is happening in the market and spending your time
> on initiatives that resonate with the market. You did alot of smart
> thinking about XRIs -- I just don't get a sense it is resonating with
> the market.
>
> -- Dick
>
> _______________________________________________
> general mailing list
> general at openid.net
> http://openid.net/mailman/listinfo/general
>



-- 
Gabe Wachob / gwachob at wachob.com \ http://blog.wachob.com

This ideas in this email: [ ] I freely license   [X] Ask first   [ ]
May be subject to patents