Re: OO Methodolgy Perspective



On Mon, 18 Feb 2008 17:01:48 +0000, H. S. Lahman wrote:


One of the problems with Structured Programming was that everything
about the solution was expressed in computing space terms. (The notion
of 'analysis' was applied to requirements; understanding what the
problem was.)

...[SNIP]....

Similarly, OOD is an elaboration of the OOA solution that addresses the
strategic needs of the computing space to address nonfunctional

...[SNIP]....

(One doesn't use OO
techniques to write a Quicksort program even if one can identify object
abstractions like PivotPoint and Partition in the algorithm.)


WOW, one of the most intuitive explanations I have ever see, though not of
my kind of thinking style.


Don't kid yourself that the OOP-based agile methods don't do OOA/D. Many
of the OOA/D construction techniques have been factored into things like
design patterns and refactoring practices. In addition, all the
OOP-based agile processes have OOA/D steps built into them (e.g., XP's
System Metaphor, CRC cards, etc.).

I know that DPs came from OO but did not know that Agile also came out
from there,that surprises me.


I haven't talked to Richard in over 40 years, but back in the '60s he
...SNIP....

He was also very much a Free Spirit;

yep, me too :)


I would be surprised if
he would embrace Structured Programming, much less a paradigm like OO
that imposes counterintuitive discipline on the development.

get surprised then. I had an email conversation with RMS where I asked him
to guide me on which part of GNU project I should work on. I said I
have only 2 skills to offer: C++ and OOA/M/D. He said *something-like*
this:

"I do not like C++ as a language, it is complex and clumsy. GNU software
do not use OO techniques. There was great noise over OOA/D in the 90s but
after some years of research I do not think it is much better. We mostly
use C. Good luck for your job search and you can't help much here"


[If he did
start using OOPLs, I'll bet the code looks just like Alan Holub's --
procedural C programs with strong typing. B-)]

:D



As far as open-source is concerned, from what I have seen of it, it is
driven more by developer culture (i.e., kindred spirits gravitate to one
another) than technical concerns.

point to be noted, my lord ;)


My
thumb-nail characterization is that people who want to be in complete
control of the machine at the code level gravitate to open-source while
those who are more process-oriented tend to migrate to more formal,
model-oriented development.

Aha... here you go...

I Love system programming. Now I can see the connection emerging, a
Patterns is becoming clearer to me.

Once, I spent 24 hours without any food at all and with only 1 glass of
water, I was sitting in front of my computer reading articles about OS
developemnt, Hurd papers and UNIX Programming and what not else. I did not
even realise that I was hungry. I never did that for C++ or OO. After 18
hours of continuous work I realised that I forgot something and after 17
more minutes it came to me that I forgot to eat my lunch and dinner :P .
What more I still did not not wat or drink anyting for next 7 or 8 hours.

So may be that's the system-programing brain that feels some anti-thinking
to OO idea.


OTOH, I think plenty of open source software is done using the OO
paradigm. In fact, I think the OOP-based agile processes are ideally
suited to open-source (provided somebody has already made the overall
systems engineering).

I agree about agile but and not about OOA/M/D model.


BTW, thanks a lot for taking out time to write sucha a lengthy and
exlanative post. It was quite helpful and will help many newbies in the
future :)


--
http://lispmachine.wordpress.com/

.



Relevant Pages

  • Re: OO Methodolgy Perspective
    ... on a methodology like XP (eXtreme Programming) or other Agile ... One of the problems with Structured Programming was that everything about the solution was expressed in computing space terms. ... One thing the OO paradigm brought to the table were a set of step stones to bridge that gap in a systematic way. ... But if the OOA/D has been done correctly, ...
    (comp.object)
  • Re: [PHP] Json.php
    ... I'm not very much into OOP, I'm very much into programming techniques ... But try a bit of functional programming in a language like PHP without ... see the win in splitting up the 'driver' and 'connection manager'. ...
    (php.general)
  • Re: Vile Hypocrasy (was: something else)
    ... >> questions his programming expertise, but he's free to call others ... > troll-baiting. ... in the way in which compiler techniques have ... trolling which are themselves trolling. ...
    (comp.programming)
  • Re: Is "Whidbey" going to de-skill developers ?
    ... wholly a good argument but cutting loose to let anybody try programming is ... because they do not learn disciplined techniques. ... Keep ASP from its last form better keep Delphi ...
    (microsoft.public.dotnet.general)
  • Re: SproutMethod.pdf
    ... >programming very troubling indeed. ... >also be seen and quantified in the design disaster of many OO ... However, I find, as I gather you do, that the applicability of the techniques ... I know how to write good programs sometimes, using mathematics, but as you point ...
    (comp.object)