Re: Holistic, Rational, Scientific Development: An Outline

From: Universe (universe_at_tAkEcovad.OuT.net)
Date: 09/30/04 

Date: Wed, 29 Sep 2004 18:33:27 -0400

Andrew McDonagh <news@andrewcdonagh.f2s.com> wrote:

> Universe wrote:
> >>>And as I see "rationalism" it's more so as I state in another post
> >>>today: the ability of proven knowledge - that is *theory* - to
> >>>rationally inform and *lead* our work practice.
> >
> >
> > I differ from Webster's in stressing that reason, theory, reasoning
> > conceptualization should play the *leading" role, in our activities
> > while practice should verify.
> >
> > Directly opposed to that, philosophical pragmatism and empiricism that
> > underlie XP places greater emphasis on the results of recent, or
> > immediate practice. It favors the aphorism, "if it works do it".

> 'Applying' the XP practise do indeed favor gaining and using what ever
> knowledge can be gained from the most recent work.

But without:
        ` first reviewing related, or contextual work for a time period
domain experts deem appropriate for the matter prior
        ` secondly doing at least minimal overall investigation of key
relevant aspects

doing whatever seems to work right now easily leads to at best the need
to reverse work and at worst serious failure.

> > Whereas I emphasize the value of studying prior, historical, or already
> > summarized experience in what we are doing and applying lessons learned
> > creatively according to specific context. This is the 1st part of what I
> > mean by "theory leads".
 
> But this sounds exactly like the approach that Beck et al took with how
> projects they had previously worked upon and resulted in what we all
> know as XP.

While some times XP verbiage talks about:
> > ... studying prior, historical, or already
> > summarized experience in what we are doing and applying lessons learned
> > creatively according to specific context

Overall, XP promotes going with whatever currently is working. Being
objective that is their *main* characteristic--look, and feel. How
often have you honestly read/heard *most* XP notables emphasize applying
the rational take. Even Beck's overall look and feel is "do it now".

> > The 2nd part of what "theory leads" is that our current activity should
> > be mapped out in always amenable theory, ideas as a plan. "Amenable"
> > plans means that they may modified, or replaced in toto during the
> > course of project IID - iterative and incremental development.
> >
> > Amenable plans should largely be constructed based upon:
> > ` ideas, theory, thinking, concepts, reasoning from prior activity in
> > contextually similar and relevant circumstances
> > ` ideas, theory, thinking, concepts, reasoning resulting from
> > investigation, and discovery of the current project context, and
> > goals.
> >
> > Plans may be modified based upon system user, management, domain expert,
> > or developer feedback from activities during the course of the project.
> > Projects should have:
> > ` a panoply of mechanisms for eliciting and incorporating feedback
> > during its IID
> > ` multiple times of regularly scheduled feedback sessions between
> > system users, management, domain experts, and or developers.
 
> Isn't that what XP strives for, but traditional aapproaches discouraged?

Yes. But this is about the whole approaches which in essence are
opposed.
 
> > A 3rd related facet of this is that the project should evaluate, weigh,
> > tradeoff and otherwise make decisions with a holistic, systems
> > perspective.
> > They should be made always after understanding and
> > considering together as whole, all key relevant parts of the whole
> > project. What is part and what is whole differs each kind of project
> > decision. But both the particular whole and parts involved should
> > identified and handled in the foregoing manner for all major project
> > decisions.

> Absolutely, even on an XP project the customer would take the same
> approach when making decisions. And importantly, it happens all the
> time, not just at the beginning or at major milestones.

XP does not stress making this kind of decision in anything, but a
secondary way in what I've read.

> > Of course this means employing a use case leading, model drive
> > architecture approach as with UP and MDA at OMG. [See www.omg.com for
> > details.]
 
> From my experience I have to disagree.

That's a fundamental difference. From industry literature and stats -
Gardner, JD Power, the journals & others - the mainstream does apply use
case led, MDA, you and XP don't.

> > High level system design should explicitly be based upon or embed an
> > object model of real actual domain entities and processes that a play
> > and role insofar as they are relevant to project use cases as whole -
> > project scope. There are a number of reasons this help create the most
> > optimal OO designs, and I will detail in soon to follow posts over the
> > next 2-3 days.
> >
> > Finally for completion the project should apply relevant contextually
> > objective facts and a scientific: hypothesis, experiment, theory back to
> > practice and so on approach.
> >
> > These are significant kind of rational holistic practices and viewpoints
> > that a truly "scientifico-engineering" approach to software development
> > takes.
> >
> > That is as opposed to the:
> > ` piecemeal
> > ` frequent stovepipe design cul-de-sac waste of resources turnaround
> > ` part above whole designs and decisions
> > ` coder-centric
> > nature of XP project development.
> >
> > Again, for those whom haven't *actually* read up on the essentials of UP
> > and MDA to best contribute to improving OO and software engineering,
> > which should be the intent, the goal of discussion on these matters. It
> > is at least as far as I am concerned.
>
> Elliot, you may have posted intending to discuss ways of improving OO
> and software engineering, but for me at least, it reads more like an UP
> & MDA vs XP rant.

A rant would not raise *concrete* *specific* major differences. Your
pro-XP stance obviously would make you sour on this.
 
> If you truly want to discuss ways of improving OO & software
> Engineering, then can we at least stop the process wars that this group
> is having....its akin to Unix vs Windows vs Mac... its also boring.

Process is at the core of improving OO & software
Engineering.

And you would have a better grasp of XP as espoused by its notables on
comp.object, by reading Google archives of *years* of debates where a
cursory perusal would inform a comp.object newbie ( < 1 year ) of the
wars revealing the true look and feel of XP--and Beck was in on it also.

Plus read all of XP/Jeffries Wiki and the other XP web places
objectively and the obviousness of how the above matches actual XP look
and feel comes across again.

Elliott 

-- 
How can Bush be winning on terrorism when:
  ~ the war in Iraq diverts military and $$$
     from Al Queda et al?
  ~ Iraq is NOW a tremendous hot bed of terrorism?
How can Bush be an expert at foreign affairs when:
  ~ he's bungling the war in Iraq as just judged BY 
	INTELLIGENCE EXPERTS?
  ~ he's alienated and worsened world US support?

Relevant Pages

  • Re: Holistic, Rational, Scientific Development: An Outline
    ... > I differ from Webster's in stressing that reason, theory, reasoning ... > while practice should verify. ... and software engineering, but for me at least, it reads more like an UP ...
    (comp.object)
  • Re: XP Requirement Analysis?
    ... what do you find circular. ... I differ from Webster's in stressing that reason, theory, reasoning ... while practice should verify. ... Amenable plans should largely be constructed based upon: ...
    (comp.object)
  • Re: Was not the Christ-Child Holy?
    ... For whatever the reason, ... reason for many deaths happening in cases when a person quits ... rather from the rush of air to the system upon the start of CPR. ... be that in the future the practice of current-day CPR will be placed ...
    (talk.religion.bahai)
  • Re: C Programming
    ... > reason for why it matters. ... There are books around that carefully do not teach bad practice, ... this book doesn't use void main. ... is int mainor int main. ...
    (Fedora)
  • Re: Circumcision
    ... >reason for it if you have phimosis or something and, ... >particular barbaric practice. ... So loudly that my wife heard ... sat in silent, wincing sympathy. ...
    (soc.culture.scottish)