The point is: is an extensive rewording compatible with document stability
(maybe also reputation)? I think we should take a decision about this
first. The next step is simply a s/resource/anything/g.
Augusto
2012/10/29 Ralf Nyren
Hi Augusto,
You really pointed out a weak spot in the spec. I have been rather unhappy with the "resource instance" terminology for quite some time (although I am probably the one to blame for it being introduced in the first place ;)
Your suggestion to put in a clarification the first time "resource instance" is mentioned is the least we can do. I will include that in the next Core draft update.
However, I have been thinking if we could try and fix this with some more clever wording. The basic problem is that we want a common name for instances of both OCCI Resource, OCCI Link and any sub-type derived from them. But _not_ an instance of Entity itself since it is an abstract class.
We could probably say "entity instance" instead of "resource instance". It would still not be optimal since you cannot instantiate OCCI Entity but at least the confusion about OCCI Link would be eliminated.
What do you all think?
regards, Ralf
Ralf,
A note about the Mixin section (sect. 4.4.4), which is fact extends to (r|R)esource definition.
Throughout the whole section the doc explains the relationship of mixins with (r)esource instances. But the term resource is overloaded: lower case and followed by {instance} it is an instance of an Entity, upper case is one of the two Entity subtypes. The reader may have the (wrong) impression that Links (the other subtype) cannot be associated with Mixins. The picture is complicated by the fact that (R)esources are defined in a following section, and forward references degrade readability. The fact that Figure 2 explains that mixins can be associated with Entities (not only Resources), and that the glossary is even more explicit about that does not help enough the first-time reader.
If you agree with me, one way to solve the problem is to state explicitly the overloading. I'd add two sentences: one the first time the term (r)esource is used, and another when the entity is defined. The sentence should be an explicit warning: "A {\em resource instance} must not be confused with the {\em Resource}, as it will be defined in section..." since the possibility of a misunderstanding is real (in my experience :-)). A pointer to the glossary can be helpful.
Thank you
Augusto
2012/10/1 Ralf Nyren
Hi,
I have updated the OCCI Core document to include the feedback on the last errata draft.
Please find a PDF attached and refer to the Git commit log [1] for details.
The document now contains an Errata Summary and is (in my view) close to its final state. If you have any pending corrections for OCCI Core now is the time to speak up!
Very important to read and provide feedback on the Core errata update and the JSON Rendering spec.
regards, Ralf
revisions/core-errata/changes/**core.tex< http://redmine.ogf.org/projects/occi-wg/repository/revisions/core-errata/cha...
On Tue, 25 Sep 2012 17:54:48 +0200, Feldhaus, Florian < florian.feldhaus@gwdg.de> wrote:
Hi,
I just spoke with Thijs and we would suggest to skip the TelCo today
and
encourage everyone to join the TelCo next week 18:00 CET. We would
On Wed, 17 Oct 2012 11:10:15 +0200, Augusto Ciuffoletti
wrote: like to discuss all remaining issues with the OCCI Core and OCCI JSON documents and then prepare them for submission to the OGF review process.
It is very important, that everyone checks the documents again. The latest versions of the documents can be found here: http://redmine.ogf.org/**projects/occi-wg/repository/**
revisions/core-errata/changes/**core.tex< http://redmine.ogf.org/projects/occi-wg/repository/revisions/core-errata/cha...
revisions/json/changes/json_**rendering.tex< http://redmine.ogf.org/projects/occi-wg/repository/revisions/json/changes/js...
Cheers, Florian
_______________________________________________ occi-wg mailing list occi-wg@ogf.org https://www.ogf.org/mailman/listinfo/occi-wg
occi-wg mailing list occi-wg@ogf.org https://www.ogf.org/mailman/listinfo/occi-wg
-- Augusto Ciuffoletti Dipartimento di Informatica Università di Pisa 56100 - Pisa (Italy)