W3C home > Mailing lists > Public > public-swbp-wg@w3.org > November 2005

Re: [SE] ODA draft

From: Phil Tetlow <philip.tetlow@uk.ibm.com>
Date: Wed, 16 Nov 2005 15:05:07 +0000
To: Jeremy Carroll <jjc@hpl.hp.com>, philip.tetlow@uk.ibm.com, pan@cs.man.ac.uk, michael.f.uschold@boeing.com, christopher.welty@us.ibm.com, adityak@wam.umd.edu, rector@cs.man.ac.uk, ewallace@cme.nist.gov, dlm@ksl.stanford.edu, dwood@tucanatech.com, Grady Booch <gbooch@us.ibm.com>, cliff.jones@newcastle.ac.uk
Cc: public-swbp-wg@w3.org
Message-ID: <OFC1D3B8AE.A4E17466-ON802570BB.00489A9F-802570BB.0052F0D8@uk.ibm.com>


Many thanks indeed for this extremely constructive contribution, your
opinions are indeed valued.

As you can see from the text below I am in personal agreement with most of
your comments, but would like to discuss further with the taskforce (week
Tue). Even so, I think that there are still some outstanding style issues.
Again we will try to discuss at the next TF telecon.

In particular Ill pass this mail on to Grady to see if I can get the
reference I mention..

Best Regards,

Philip Tetlow
Senior Consultant (Certified Technical Architect)
IBM Business Consulting Services

Mail: IBM United Kingdom Limited, 1175 Century Way, Thorpe Park, Colton,
Leeds, LS15 8ZB
Current Assignment: DWP BPRP (Metadata)
Mobile: +44 (0)7740 923328
Email: philip.tetlow@uk.ibm.com

             Jeremy Carroll                                                
             Sent by:                                                   To 
             public-swbp-wg-re         public-swbp-wg@w3.org               
             quest@w3.org                                               cc 
             15/11/2005 17:39          [SE] ODA draft                      

Comments on ODA draft

1) abstract is not an abstract. An abstract should be a short summary of
the content of the paper.
The end of the introduction is fairly close to an abstract

How about: This note outlines the benefits of applying knowledge
representation languages common to the Semantic Web, such as RDF and OWL,
in Systems and Software Engineering practices. It is primarily written from
a Systems and Software Engineering (SSE) perspective.

2) Suggest moving the definition of acronym (SSE) from abstract into its
first use in main text (currently in Target Audience)


3) I found the history section 2.1, 2.2 off-putting.
   This is largely stylistic. For example, I was happy with the
treatment of the same history in one of the presentations at the SE
Workshop in ISWC. I suggest taking some of the treatement from Rudi
Studer's slides.

My main stylistic problem is that there is metanarrative concerning
progress of SE.

Metanarratives were originally introduced in religions, particularly
Judaism and its derivatives - and place the present within a grand
picture, typically leading from some problem state to a wonderful solution.

In the 19th and 20th centuries metanarratives concerning progress were
very popular, for example, communism.

However, since the second world war metanarratives have been falling
into disrepute, partly due to some obvious abuse of them by various

In terms of your text, the migration from functions to subroutines and
libraries, to objects to bundles etc is given without any hint of doubt
or distancing.
For example "such approaches could no longer effectively manage shared
data and concurrency" may or may not be true; a more limited statement,
which does not commit so much to the present being an improvement on the
past, rather than merely a variant, would be say:

"'object oriented' programming was introduced in the mid 1980s with the
hope of better managing shared data and concurrency"

This does not commit as to whether that hope has been fulfilled, or even
as to whether or not there was a problem. This phrasing hence makes
fewer assumptions that the reader and writer of the document share
judgements about what is good and bad in system design, and is more

My personal opinion is that I agree strongly with your position (and love
your historical support), but the task force's consensus was that a major
audience for this note would have no significant background in SSE
(undergrad students for example) and would hence value a ‘gentle’
introduction to this area. I further agree with your point relating to
‘doubt or distancing’, but I think that the task force should debate you
comment further – The original text in this section did not come from me
but I do not believe that the motivation behind this was blatant
self-promotion for SSE’s supposedly flawless past.

I have particular problems with the use of the word 'logical'. Typically
metanarratives have a logic of their own which is often evil, which is
one of the key reasons they have been falling into disrepute. (I am
unclear as to whether I am writing a metanarrative of metanarratives)

Strongly Agree: Changed to "Today we are just starting to see further steps
in software systems' construction"

4) Another stylistic problem was the use of unattributed positions.
e.g. "Indeed many experts now consider " - reference please. Or rephrase
to make a statement which you wholly own, and don't step away from.
Again "Similar to current work in Semantic Web Services" which work?
"practitioners" who?
"a commonly asked question" who is asking?

Agree: "Indeed many experts consider" changed to "Indeed one could now
Agree: "Similar to current work in Semantic Web Services" changed to "The
aim here is to make such rules unambigous"
Mildly Disagree: Surely the use of a deliberately generalised term such as
"practitioners" is acceptable in this context to appeal to as wider
audience as possible. Neverthless changed to "SSE
practitioner"...practitioners also changed to "industrial professional" in
the Target Audience section...hope that's better?

5) I've scribbled "speculation" beside the subbullets under mechanism in
sectioon 3.1, I think particularly triggered by "could be viewed as".

Agree: Fair comment - wording changed to be more definite (i.e. "can"
rather than "could"), but dont think there is much debate about the
classification "types". The names of the classifications may me a little
different but content can only ever comprise data, function or a
combination of both. Nevertheless if you have better suggestions for this
section, or you would prefer it taken out all together...just let us know.

6) Again phrasing under "a collection of Semantic Webs"
I think the position stated is defensible but the language used could be

Agreed: Changed further to valued input from Jim Hendler

7) Notes should not promise or suggest much further work e.g "a
potential future note form [sic] this taskforce"

Agree: Removed

8) Again the "speculative" word along side the bullet points after:
"employing such techniques:"

Not sure - further comment from you required:The use of composite keys is
proven best practice in the RDB world and everyone knows that it can be
applied to RDF and OWL, we just do not have convetion on practice yet. Does
lack of convention preclude mention of such practices?

9) I found the references odd - why divide into normative and
informative, since this note has no normative content.

;0): Actually we did this following your earlier advice - sorry!

10) The phrase "notoriously ambiguous UML" is again unattributed and
unnecessary, except to persuade the reader by rhetoric of a position
that may be defensible, but is not defended in this document. Either
back up such an assertion with references and/or argument or delete it.
It is unnecessary.

Agree (I think): Here I was refering to conversations with guy's like
Grady. Ill ask him directly if he is in print on this matter and then add a
reference if applicable

I need to review the proceedings of

once I have I'll try and send more comments.
Overall, as you know, I enjoyed the workshop immensely; and both there
and in the primer, there seems to be plenty of better grounded material
to make a compelling case for many of the positions taken in this paper.


Received on Wednesday, 16 November 2005 15:05:51 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:09:45 UTC