Re: [oslc-core] Review comments on Attachments 3.0 draft

[Samuel Padgett <spadgett@us.ibm.com>]

Ian writes:

> > I'd say no. An LDP 2.0 server might not be 1.0 compliant.
> > The intent is we are compliant with some version of LDP, 1.0 or a future
> > version.
>
> But you do not know that compliance with a future version of LDP
> will be consistent with Attachment 3.0.
>
> Eg if LDP 2.0 is not backwards compatible with 1.0, how can you
> state in the Attachments 3.0 spec that 2.0 would be acceptable (as
> you intend by writing "at least")?

Understood. Let's discuss on the next call? I don't have a strong
opinion on this, just clarifying what "at least" meant in the spec.

> Removal of an attachment (whatever ways there are) should be those
> defined by LDP.  I see no reason for there to be Attachment-specific
> normative statements since these are covered in LDP/HTTP.  Suggest
> we clarify LDP spec. then come to decision.

Thinking more about it, I'm fine with that. We do have an example as
well later on.

> > You make a good point. We have a problem in that the attachment is a
> > non-RDF source, so there's no shape. I'm open to suggestions. We could
> > remove discussion of wdrs:describedBy from this section.
>
> I don't think shape is useful here.  A server choosing to create an
> attachment descriptor MUST link to it from the attachment.  An
> attachment MUST NOT link to more than one descriptor.

Are you saying to add additional requirements on this? I'm leaning
towards just removing wdrs:describedBy from section 9 and perhaps using
it in the examples only.

--
Samuel Padgett | IBM Rational | spadgett@us.ibm.com
Eclipse Lyo: Enabling tool integration with OSLC