Question about entityID

[Ken Weiss]

That's very clear and complete, Tom. Thanks!

--Ken


------------------------------------------------------------
Ken Weiss                                 ken.weiss at ucop.edu
UC Office of the President              510-587-6311 (office)
California Digital Library              916-905-6933 (mobile)
UC Curation Center
415 20th Street, 4th Floor
Oakland, CA 94612






On 7/2/13 2:04 PM, "Tom Scavo" <trscavo at gmail.com> wrote:

>On Tue, Jul 2, 2013 at 4:24 PM, Ken Weiss <ken.weiss at ucop.edu> wrote:
>>
>> The problem is that I created an entityID called
>> dmp2-development.cdlib.org.
>
>I think you mean an entityID with host part "dmp2-development.cdlib.org".
>
>> This is NOT a host name. There is not, and
>> never will be, a server that responds to that name. There will never be
>>an
>> entry in any DNS system mapping that name to an IP address.
>
>That's fine.
>
>> However, Internet2 won't register our metadata until the administrative
>> contact for the .cdlib.org domain approves it. This is a problem because
>> the email address on our registration is out of date, and appears to now
>> route to /dev/null. It doesn't bounce, but it doesn't seem to land in
>> anyone's inbox, either. I know that's not a good thing, and we're fixing
>> it.
>
>That's good. Once it's fixed, I'm sure your metadata registrar will be
>happy to resend the email ;-)
>
>> In the meantime, is there any reason I couldn't make my entityID
>> dmp2-development, or dmp2-development-CDL, or
>> kens-terrific-shibboleth-enabled-application?
>
>Well, an entityID MUST be an URI, so as long as that basic requirement
>is met, there is hope.
>
>> The entityID is simply a
>> unique identifier that is used as a primary key to retrieve my SP's
>> metadata from an IDP.
>
>Functionally, yes.
>
>> It has no other purpose, and could even be an opaque
>> unique identifier with no semantic value whatsoever, like
>> 4356.5341.3566.5657, right?
>
>No, I'm afraid not. First, an entityID MUST be an URI as mentioned
>above. Second, it MUST satisfy InCommon's requirements for an
>acceptable entityID (https://spaces.internet2.edu/x/eAUjAQ). The
>requirements listed may seem arbitrary, but they are not. Being names,
>some entityIDs are definitely better than others. The conventions
>documented in the wiki are known to give rise to "good" entityIDs.
>
>The entityID is probably the single most important piece of metadata.
>Choose it well, and choose it to last. Changing the name of a
>production SAML entity may be painful.
>
>Tom
>--
>To unsubscribe from this list send an email to
>users-unsubscribe at shibboleth.net