RE: [soa-rm-editors] FW: [chairs] Draft ASIS under review: mandatory policy? Please review and comment by 1 March

["Peter F Brown" <peter@justbrown.net>]

Ken:
spot on with the question.

I was involved in an earlier review cycle, and even discussed parts of it
with Duane in Vancouver last year: I had thought that my views were a bit
odd-ball until speaking with Duane and, plucking up courage and talking with
more people, realised soon that there were very few people that supported
the proposals.

My main objection has been, simply stated, that the effort to create a
naming scheme based on (some of the)semantics of the artifact always
outweighs the benefits.

I do think that it is worth replying, and I will be making a submission: it
is interesting that Robin has written offlist to a number of people,
including myself (maybe you too Duane?), asking that they respond,
suggesting a certain nervousness on his part, or others, regarding its
implementability.

In summary, while I welcome the move towards a mandatory set of metadata
(caveat: the bloody things have a habit of changing over time), I do not
accept that that such metadata should determine the naming of an artifact.

More to the point, I think the naming convention will be top heavy and
ultimately confusing and unworkable, with a resultant soup of samizdat
approaches by more practical minded TCs...

In more "information architecture" terms: metadata represents properties of
a relationship between two artifacts (often one abstract and another
concrete), and not properties of the artifact itself.

For example: we will often casually state that document x is version 1;
whereas the correct statement would be that document x (concrete artifact)
represents version 1 of a particular work y (abstract artifact).

RDF triples and Topic Maps handle these concepts very well.

The important point to note is that often the *artifact* does not change,
whereas the nature and properties of its relationship with other artifacts
does: a single document can change "status" from a draft to a final document
without a byte of data changing. Equally an "abstract artifact", say a
committee specification can have associations with several documents, each
representing a different property or set of properties (e.g. draft, final,
in English, in PDF, in unofficial translation in Chinese in ODF, etc).

On my own website I've eliminated semantics completely from filenaming
conventions, and merely have two series of numbers to represent "object
concepts" (this presentation, that article, my resume), the URLs of which
are never changed once attributed, and actual content pages (when content
changes, it is a new file, and given a new filename). What changes is the
relationship, and thus the "triple". This means that as long as links are
only ever made to the concept URLs, the user is always directed to the most
recent content page, and there is no such thing as a dead link...I digress:
I will be presenting these sorts of arguments to TAB in the hope tat they
will end their folly...

On a less kind note: where are the information scientists or trained
librarians in the group that cooked up this draft?

Peter

-----Original Message-----
From: Ken Laskey [mailto:klaskey@mitre.org] 
Sent: 20 February 2006 21:12
To: Duane Nickull
Cc: soa-rm-editors@lists.oasis-open.org
Subject: Re: [soa-rm-editors] FW: [chairs] Draft ASIS under review:
mandatory policy? Please review and comment by 1 March

Has anyone been involved in the process or reviewed previous versions?   
Is this routine stuff or does it have the potential to be annoying but not
commensurately productive overhead in the future?

Ken

On Feb 20, 2006, at 3:07 PM, Duane Nickull wrote:

> We need to respond.
>
> D
>
> *******************************
> Adobe Systems, Inc. - http://www.adobe.com Vice Chair - UN/CEFACT  
> http://www.uncefact.org/ Chair - OASIS SOA Reference Model Technical 
> Committee Personal Blog - http://technoracle.blogspot.com/
> *******************************
>
>
> -----Original Message-----
> From: James Bryce Clark [mailto:jamie.clark@oasis-open.org]
> Sent: Friday, February 17, 2006 4:11 PM
> To: chairs@lists.oasis-open.org
> Subject: [chairs] Draft ASIS under review: mandatory policy? Please 
> review and comment by 1 March
>
> TC leadership:
>
>      Please take a careful look at the current member review draft of 
> the "OASIS Standard Artifact Identification Scheme (ASIS)" at [1], and 
> ask your editors and other interested TC members to do the same.  Our 
> Board of Directors plans to consider in March whether to make it, or 
> portions of it, mandatory policy after the current review is 
> completed.
>
>      Previous versions of this document were called "Artifact 
> Identification Requirements" and "Artifact Naming Guidelines", and 
> were the subject of an earlier review in June 2005 [2].  This proposed 
> policy would provide the technical details behind the "file naming 
> scheme" rule described in Section 2.18 of the TC Process:
>
>> "... All documents and other files produced by the TC, including 
>> specifications at any level of approval, must use the OASIS file 
>> naming scheme, and must include the OASIS copyright notice. ..."
>
> The Board plans to take into account the input from this second 
> review, and any recommendations the TAB makes as a result, when it 
> considers the document in March.
>
>      The document itself, and a disposition log of comments from the
> 2005
> review, can be found at the links listed in [1].  Please provide your 
> feedback between now and the deadline for comments, 1 March 2006.
>
>      Thanks and best regards  JBC
>
> [1]
> http://lists.oasis-open.org/archives/oasis-member-discuss/200602/
> msg00000.ht
> ml
> [2]
> http://lists.oasis-open.org/archives/tc-announce/200507/msg00001.html
>
>
>

------------------------------------------------------------------------
------------------
Ken Laskey
MITRE Corporation, M/S H305     phone:  703-983-7934
7515 Colshire Drive                        fax:        703-983-1379
McLean VA 22102-7508