[security-dev] REST API for user provisioning

[Anil Saldhana]

On 03/11/2013 08:00 AM, Pedro Igor Silva wrote:
> Maybe we can review the Attribute API and have something like that:
>
> User user = // create user
>
> user.setAttribute(Attribute.create("profileUrl", "http://company/users/profile?id=121")); // the same as new Attribute("",""), just a helper method
>
> Attribute addresses = Attribute
>
>      .create("addresses") // parent, no value defined
>
>          .add("address1") // child of "addresses" and also a parent
>              .add("postalCode", "123")
>              .add("streetAddress", "123")
>              .add("multivalued", "1","2") // multi-valued attribute for "address1"
>
>          .add("address2")
>              .add("postalCode", "456")
>              .add("streetAddress", "456")
>
> user.setAttribute(addresses);
>
> Another thing we can review is the getEmail method on the User interface. Maybe we should support a String array to allow 1+ emails for users.
Good question.  Users typically have a primary email and a secondary 
email. They may also qualify the email addresses such as work email, 
personal email etc. Is it better to use the attributes in this case?

>
> Regards.
> Pedro Igor
>
> ----- Original Message -----
> From: "Shane Bryzak" <sbryzak at redhat.com>
> To: security-dev at lists.jboss.org
> Sent: Friday, March 8, 2013 7:58:31 PM
> Subject: Re: [security-dev] REST API for user provisioning
>
> Would we be targeting SCIM 1.1 or 2.0? Also, it looks like we can
> already support most of this already with our identity model, although
> for complex types it might be a little bit of a challenge.  I guess one
> possibility though is to just make the Attribute value an array of
> Attributes itself.
>
> On 09/03/13 06:53, Anil Saldhana wrote:
>> Hi All,
>>      now that we have an excellent IDM subsystem as part of PicketLink3,
>> we need to next look at incorporating SCIM
>> (http://www.simplecloud.info/), a set of standards surrounding REST API
>> for Cloud Provisioning. SCIM is part of the IETF.
>>
>> Probably in the PicketLink 3.1+ timeframe.
>>
>> Regards,
>> Anil