[OpenID] SREG 1.x attributes

[Peter Williams]

I think we need to decide where both sreg and ax play, in the wider eco space. Both leverage the openid association for data origin auth, of course (as will nat's agreement protocols).

Form filling (sreg) is not the same as interacting with an attribute authority (ax) -as part of a web sso evet, or statelessly.

I already have 15 form filling products in my environment. One more (sreg) will make no difference. 100 more variants of sreg is a different matter. Vendors will inevitably just piss on each other over profile support, claiming that their sreg profile is more suitable than the other for some, per-vertical form filling need, leading to proliferation - for little value.

________________________________
From: Shane B Weeden <sweeden at au1.ibm.com>
Sent: Sunday, November 30, 2008 2:15 PM
To: david at sixapart.com <david at sixapart.com>
Cc: general at openid.net <general at openid.net>
Subject: Re: [OpenID] SREG 1.x attributes


I don't have any problem with the development and promotion of AX. The point was simply that we have an SREG 1.1 spec in draft, and I can't see any good reason for it to include in the 6th paragraph of section 4 ", and all included fields MUST be taken from the set of fields defined in this specification." It appears to be boxing something in that simply doesn't require a box. There are plenty of intranet or 1:1 partner scenarios where OpenID with flexible SREG is an equally viable, and arguably simpler solution.

Regards,
Shane.




David Recordon <drecordon at sixapart.com>
Sent by: general-bounces at openid.net

01/12/2008 07:29 AM
Please respond to
david at sixapart.com




To
        Peter Williams <pwilliams at rapattoni.com>, "Eddy Nigg (StartCom Ltd.)" <eddy_nigg at startcom.org>, "general at openid.net >> general at openid.net" <general at openid.net>
cc

Subject
        Re: [OpenID] SREG 1.x attributes







I think it's far more of a timing thing.  SREG has been supported by
the majority of smaller OpenID Providers and thus some Relying
Parties.  The large Providers (AOL and Yahoo!) didn't provide support
for either spec.  Google now supports AX and MySpace is planning to as
well.  AX will overtime become prevalent as more OPs provide data via
it.

This is why I believe that we should fix SREG to work with 2.0 though
focus effort on adoption of AX.

--David

On Nov 30, 2008, at 1:13 PM, Peter Williams wrote:

> I suspect ax is not too popular as it competes with ldap/ldif - and
> has no real reason to exist. The pain of making yet another
> attribute schema (for what?) is high.
>
> Best if one extends sreg - one field that can deliver an lidf
> stream. Ldif is the sytax, and existing ldap schemas can be used.
> Extensibility is then handled within the ldap world.
>
> -----Original Message-----
> From: Eddy Nigg (StartCom Ltd.) <eddy_nigg at startcom.org>
> Sent: Sunday, November 30, 2008 1:09 PM
> To: general at openid.net >> general at openid.net <general at openid.net>
> Subject: Re: [OpenID] SREG 1.x attributes
>
> On 11/30/2008 12:39 AM, Martin Atkins:
> Eddy Nigg (StartCom Ltd.) wrote:
>
> Which is probably because AX was supposed to take care of it. However
> SREG is widely implemented and I agree that we should make it possible
> to easily extend SREG.
>
>
> If you make it possible to extend SREG, what you end up with is AX
> with
> a few predefined attributes baked into the spec and no ability for
> those
> attribute to be updated by third-parties.
>
> I'm not sure anyone actually uses the ability for the attributes to be
> updated by third-parties, so for all intents and purposes you'd just
> be
> reinventing AX as far as I can tell.
>
> Yes, I mentioned that AX is supposed to take care of this.
>
> What do you see as the value of having two parallel specs that both
> have
> the same functionality apart from one having a feature that no-one
> uses?
> (or are you proposing that SREG would replace AX altogether?)
>
> But one might ask, why SREG has become popular (if we can speak
> about popularity in terms of OpenID ;-) ), whereas AX isn't widely
> implemented as far as I can tell. Maybe it's because of the pre-
> defined fields or because of its simplicity, I don't know. But
> something tells me that SREG has been successful, why not extend it
> in some form.
>
>
> Regards
>
> Signer:
> Eddy Nigg, StartCom Ltd.
> Jabber:
> startcom at startcom.org
> Blog:
> Join the Revolution!
> Phon
>
> [The entire original message is not included]
> _______________________________________________
> general mailing list
> general at openid.net
> http://openid.net/mailman/listinfo/general


_______________________________________________
general mailing list
general at openid.net
http://openid.net/mailman/listinfo/general