Re: [regrep] RIM 3.0 Comments (Cure for Cabin Fever?)

[Farrukh Najmi <Farrukh.Najmi@Sun.COM>]

It is truly great to see such a solid effort by TC members to get the 
final push to get this 3.0 baby out. Please see comments inline below.

Chiusano Joseph wrote:

> Here are some comments on RIM 3.0, version "Working Draft 01, 04 
> January, 2005" (which I believe is still the
>
> latest version that has been distributed to the TC). Apologies if any 
> have already been offered by others. I've also
>
> placed a general category in ()'s after each line, for prioritization 
> - so items that have the word "meaning" or
>
> "accuracy" in their category may be considered higher priority than - 
> for example - those labeled "Readbility":
>
> - Line 363 (Readability): Expand SAML V2.0 “ECP” abbreviation
>
+1. Fixed in draft 02

> - Line 363 (Readability): Expand Liberty Alliance “PAOS” abbreviation
>
PAOS is reverse of SOAP (cute huh). Added "(reverse SOAP)" in draft 02

> - Line 367 (Grammar): “Instances of such content…” – change to 
> “Instances of such content are referred to as RepositoryItems"
>
+1. Fixed in draft 02

> - Line 369 (Concept meaning): “Standardized metadata” – do we mean the 
> metadata MODEL itself, or the contents? I believe we mean the model, 
> which is the RIM. The contents (e.g. format of RegistryObject 
> description values) may or may not be standardized, depending on an 
> organization’s usage. Applies also to line 384.
>
We mean the model itself. The metadata is standardized by the classes 
and attributes defined by RIM. No change suggested here.

> - Line 370 (Grammar): “Instances of such metdata…” – change to 
> “Instances of such metadata are referred to as RegistryObjects"
>
+1. Fixed in draft 02

> - Line 372 (Spelling): Metaphore is misspelled
>
+1. Fixed in draft 02

> - Line 388 (Grammar): “within a canonical” – remove “a”
>
+1. Fixed in draft 02

> - Line 396 (Readability): Recommend considering adding a short 
> description of each classification scheme (e.g. “Defines normative 
> Association values…”). Also, shouldn’t this table be labeled?
>
+1. Fixed in draft 02

> - Line 406 (Readability): Recommend citing one example here of a 
> composed object (e.g. a Classification instance)
>
Added an entire section based on Goran's comment that normatively 
itemizes every composed object and defines the term in darft 02.

> - Line 415 (Readability): Recommend adding a short reference 
> to/description of the Identifiable class prior to Figure 2, since it 
> is at the top (even though there is a description in the box to the 
> left of it in the figure), the significance of it being where it is, 
> and why some classes in Figure 1 do not appear here (i.e. why they do 
> not need to be identified via an id attribute – ex: PostalAddress)
>
+1. Fixed in draft 02 by adding a small section describing this class 
just before Figure 2 as follows:

1.7.2.1 Class Identifiable

The RegistryObject class and some other classes in RIM are derived from 
a class called /Identifiable/. This class provides the ability to 
identify objects by an id attribute and also provides attribute 
extensibility by allowing dynamic, instance-specific attributes called 
Slots.


> - Figure 2 (Accuracy): Box to left of RegistryObject states that 
> “classes in the model below are referred to as core information model 
> classes”, yet the indication of such classes is only on the right of 
> that row (under ExternalIdentifier and ExtrinsicObject).
>
+1. Fixed in draft 02

> - Figure 2 (Accuracy): Shouldn’t this figure also contain the Access 
> Control Information Model?
>
No the Access Control Information Model does not define any new classes 
in RIM. It simply uses existing classes in a stylized way.

> - Line 430 (Grammar): “The remainder of this document will be 
> describing…” – change to “The remainder of this document will describe...”
>
+1. Fixed in draft 02

> Kind Regards,
> Joseph Chiusano
> Booz Allen Hamilton
> Strategy and Technology Consultants to the World



-- 
Regards,
Farrukh