Hi Sanjay,
Yes, the model is intended to help the reader understand
where the specification is in its development cycle. Merely incrementing WD loses
the relationship to prior CDs and CSs. This is the model that is described in
the rules document and to be followed.
I’m trying to work on some quick, downloadable mini-tutorials
that will hopefully make many of the fine points in the various rules and
procedures governing TCs easier to understand; this is something I’ll
definitely have to add to that list. One of the difficulties is that with
60-odd technical committees, I typically only *see* documents when they’re
at one of the approval stages so chances are there are a number of TCs who aren’t
getting it quite right. It’s my hope that at least new TCs will start off
on the right foot.
Regards,
Mary
From: Patil, Sanjay
[mailto:sanjay.patil@sap.com]
Sent: Tuesday, October 23, 2007 7:00 PM
To: mary.mcrae@oasis-open.org; Bryan Aupperle; Martin Chapman
Cc: opencsa-ms@lists.oasis-open.org
Subject: RE: [opencsa-ms] OpenSCA TC file names and URL structure
Hi Mary,
Upon reading the guideline again, I am now realizing that my
understanding of the scope of WD stage and the use of revision numbering
is different from what is described in the guideline.
I think the guideline assumes that the scope of WD expires
after the first CD is issued, and subsequent versions are
tracked as revisions of CDs, etc. So as per the guideline, an example of the
series of drafts would be as follows:
WD 01
WD 02
CD 01
CD 01 Rev 01
CD 01 Rev 02
CD 02
CD 02 Rev 01
CD 03
CS 01
My understanding however is that the WD is not just the very first
stage that terminates after the first CD is issued, but it continues
throughout the TC's life cycle. I had thought that WD is merely a convention to
identify editors' drafts. I assumed that the editors would use a monotonically
increasing sequencing numbers for tracking the WDs, and periodically the TC may
decide to vote upon one of the WDs as a CD (and subsequently CS, etc). So the
series of drafts would look something like:
WD 01
WD 02
WD 03 --> CD 01 Rev 01 -> CD 01
Rev 02 ---(Vote)--> CD 01
WD 04
WD 05
WD 06 --> CD 02 Rev 01 --(Vote)--> CD 02
--(Vote)--> CS 01
WD 07
....
Note that in the above example the WD sequencing continues
well beyond the first CD vote, and during the TC's life cycle, statements
could be made such as - "WD 03 was voted upon as CD01", "WD 06
was voted upon as CS 01", etc. In this model, the revision numbers
are used to track merely the delta changes to accommodate any minor updates
(fixing typos, etc) in finalizing that CD. After the CD vote, the editors
would fold in the changes made by the CD revisions back into the next WD and
continue to produce further WDs.
Obviously, I think that the model as I had understood is
better :-). It also happens to be successfully used by a OASIS TC I recently
chaired.
Comments? If you think that I am reading the guideline (and the
examples in there) more strictly than what was intended, and the model I laid
out above is not illegal as such, then I would be quite happy
to know that as well.
-- Sanjay
From: Mary McRae [mailto:marypmcrae@gmail.com] On
Behalf Of Mary McRae
Sent: Tuesday, Oct 23, 2007 15:08 PM
To: Patil, Sanjay; 'Bryan Aupperle'; 'Martin Chapman'
Cc: opencsa-ms@lists.oasis-open.org
Subject: RE: [opencsa-ms] OpenSCA TC file names and URL structure
Hi everyone,
OASIS has an established procedure for uniquely
identifying specifications. While the TC is free to name the actual document
files as they see fit, the formal identification process must be followed on
the actual specification documents. Details can be found here:
http://docs.oasis-open.org/specGuidelines/namingGuidelines/metadata.html#stage
and here:
http://docs.oasis-open.org/specGuidelines/namingGuidelines/metadata.html#revision
Regards,
Mary
From: Patil, Sanjay
[mailto:sanjay.patil@sap.com]
Sent: Tuesday, October 23, 2007 3:18 PM
To: Bryan Aupperle; Martin Chapman
Cc: opencsa-ms@lists.oasis-open.org
Subject: RE: [opencsa-ms] OpenSCA TC file names and URL structure
For Working Drafts, I think we can avoid revision numbers since the
seqnum itself can be incremented. In my experience, the rev nums are handy for
CDs and above, for example, while producing CD 1, the candidate drafts may go
through a few revisions for applying pro-forma changes, applying any bits that
were previously missed, etc. We could perhaps use alphabets for identifying
revision numbers in order to improve readability e.g. CD 1 Rev A, CD 1 Rev B,
...
-- Sanjay
From: Bryan Aupperle [mailto:aupperle@us.ibm.com]
Sent: Tuesday, Oct 23, 2007 11:55 AM
To: Martin Chapman
Cc: opencsa-ms@lists.oasis-open.org; Patil, Sanjay
Subject: RE: [opencsa-ms] OpenSCA TC file names and URL structure
The revnum is
for consistency with the OASIS guidelines (http://docs.oasis-open.org/specGuidelines/namingGuidelines/metadata03.html#revision)
If we can indicate the revnum as part of the seqnum, then I'm OK with
that.
Bryan Aupperle, Ph.D.
STSM, WebSphere Enterprise Platform Software Solution Architect
Master Inventor
Research Triangle Park, NC
+1 919-254-7508 (T/L 444-7508)
Internet Address: aupperle@us.ibm.com
|
"Martin
Chapman" <martin.chapman@oracle.com>
10/23/2007
02:46 PM
|
|
To
|
Bryan
Aupperle/Raleigh/IBM@IBMUS, <opencsa-ms@lists.oasis-open.org>
|
|
cc
|
"'Patil,
Sanjay'" <sanjay.patil@sap.com>
|
|
Subject
|
RE:
[opencsa-ms] OpenSCA TC file names and URL structure
|
|
For specs,
why do we need seqnum and revnum, surely we can just increment the seqnum.
Martin.
-----Original
Message-----
From: Bryan Aupperle [mailto:aupperle@us.ibm.com]
Sent: Tuesday, October 23, 2007 7:08 PM
To: opencsa-ms@lists.oasis-open.org
Cc: Patil, Sanjay
Subject: [opencsa-ms] OpenSCA TC file names and URL structure
Attached is an enhanced proposal, from Sanjay and myself, covering file name
construction for files produced by the OpenCSA TCs and the URL structure used
for these files in the OASIS document repository.
Bryan Aupperle, Ph.D.
STSM, WebSphere Enterprise Platform Software Solution Architect
Master Inventor
Research Triangle Park, NC
+1 919-254-7508 (T/L 444-7508)
Internet Address: aupperle@us.ibm.com
|