[oslc-core] Questions about the creator property

[Arthur Ryman]

Dave,

I think dc:creator should identify the user that created the resource. In 
general, services will require a user to authenticate before they can 
create resources (e.g. to prevent malicious use). The service will 
therefore have some type of user identify available at resource creation 
time, and in general that will be some type of resource. Therefore, a 
simple string is not suitable. Using a foaf:Person provides a lot of 
flexibility since it can have additional properties such user id's, names, 
etc.

I think the simplest way to provide the user identity is via the 
foaf:account property of foaf:Person, e.g. if I created a page on the OSLC 
wiki:

<dc:creator>
        <foaf:Person>
                <foaf:name>Arthur Ryman</foaf:name>
                <foaf:account rdf:resource="
http://open-services.net/bin/view/Main/ArthurRyman" />
        </foaf:Person>
</dc:creator>

Here I am simply referring to the account resource. It could have type 
foaf:OnlineAccount, but we don't have to inline that. 

Regards, 
___________________________________________________________________________ 

Arthur Ryman, PhD, DE


Chief Architect, Project and Portfolio Management

IBM Software, Rational

Markham, ON, Canada | Office: 905-413-3077, Cell: 416-939-5063
Twitter | Facebook | YouTube







From:
Dave <snoopdave at gmail.com>
To:
Nick Edgar/Ottawa/IBM at IBMCA
Cc:
oslc-core at open-services.net
Date:
05/06/2010 09:48 AM
Subject:
Re: [oslc-core] Questions about the creator property
Sent by:
oslc-core-bounces at open-services.net



Hi Nick,

Thanks for the feedback. I think we may need some tweaks in this area.

I just added some text to the Core spec to clarify how foaf:Person is
supposed to be used:
- dc:creator MUST be either a Resource or Local Inline Resource of
type foaf:Person
- foaf:Person MUST include either a foaf:name or both foaf:givenName
and foaf:familyName

That's is how things stand today. We only allow foaf:Person for
dc:creator and we only support those three foaf:Person properties. In
this area, OSLC Core was trying to approximate what was done in
earlier OSLC specs.

Of course, we still have time to change, and now is the time to get this 
right.

Seems like having some form of user ID would be a very good thing and
so I wonder:

1) Have we had user ID in previous OSLC specs and/or have we had some
discussions about the implications of specifying user ID's in
representations?

2) Any opinions about how we use dc:creator in OSLC? Should it be a
simple "full-name" string as the FOAF guidance suggests?

3) Any opinions about if or how we should specify user ID? I think
"foaf:OnlineAccount" would probably be preferable to "oslc:userID" and
preferable to pulling in the SIOC namespace.

- Dave



On Wed, May 5, 2010 at 12:32 PM, Nick Edgar <Nick_Edgar at ca.ibm.com> wrote:
> Hi all,
>
> While working on the OSLC Automation provider prototype implementation 
for
> RTC Build, some questions have come up over the exact representation to 
use
> for the dc:creator property.
>
> The core spec says that the dc:creator is a foaf:Person, but later in 
the
> Common Resource Definitions section it says "The Person resource can be 
used
> as the value for a dc:creator  or dc:contributor property. ", which is a 
bit
> ambiguous.
> Is it valid to use a simple string value instead of a foaf:Person?
>
> Having a foaf:Person value for a dc:creator property seems at odds with 
the
> FOAF spec, where they suggest keeping dc:creator as the simple string 
and
> using foaf:maker for the Person details, and that dc:creator is the same 
as
> the maker's foaf:name.
> See http://xmlns.com/foaf/spec/#term_nick
> There are some good arguments for that approach in the linked 
discussion:
> http://wiki.foaf-project.org/w/UsingDublinCoreCreator
>
> The original problem that led to me reading up on this was whether and 
how
> to include the Jazz user id (in addition to the real name).  I 
originally
> thought of just tacking on a dc:identifier property inside the 
foaf:Person,
> but foaf:nick is probably more appropriate, since the user id isn't a 
unique
> identifier for the actual person, and the scope of the user id isn't 
given.
>  Could maybe use foaf:OnlineAccount instead, or in addition to 
foaf:nick,
> but it's marked as 'unstable'; it would also be pretty verbose.
>
> Are there more specific recommendations for which of the foaf properties
> should be given? All the ones mentioned by the core spec (foaf:name,
> foaf:givenName and foaf:familyName) are listed as optional.  Is it OK to
> just give foaf:name, or if one is given should they all be?
>
> Thanks,
> Nick Edgar
> RTC Build component lead
>
> _______________________________________________
> Oslc-Core mailing list
> Oslc-Core at open-services.net
> http://open-services.net/mailman/listinfo/oslc-core_open-services.net
>
>

_______________________________________________
Oslc-Core mailing list
Oslc-Core at open-services.net
http://open-services.net/mailman/listinfo/oslc-core_open-services.net