Re: OO and Procedural Work Units (was switch string) -- JAVA400-L


On 03/07/2001 at 01:14:20 PM, owner-java400-l@midrange.com wrote:
OO programming it will take 1000 work units to initially create, and then
each update requires 25 work units.

Procedural programming will take 350 work units to initially create, and
each update will take 48 work units.

The further you take OO, the more work units I can see it taking to produce
an end product.  And, assuming it's done correctly (you don't paint
yourself
into a corner sub/superclassing everything) it will help similar
developement as well as updates (ie tying it in with inventory).

I'd love to hear other's opinions on this, especially anyone who's done
production Java and RPG/ILE (real ILE) programming on similar systems.
--- end of excerpt ---

I completely agree in concept and I like your 'work units' idea.

I've personally found the delta for both understanding
a product and making updates to a product
OO model versus a procedural model to be MUCH more
significant (probably as much or more so as the delta
between design in the models.).

Its really REMARKABLY EASY for me to load thousands of
classes/files into my tagging editor, spend
an hour or two looking at the object relationships and have
a really good understanding of the overall infrastructure
of the product. (I've actually done this for Websphere
on AS/400 even though I work on the JDBC team).

In a large procedural product, that effort is much more
difficult and leads to much higher cost to understanding
the product and to making (safe and correct) changes.

===
So, as a swag at numbers, I'd probably change the update
to be 25 work units for OO and 65 for procedural.
Between 2 and 3 times as much to
1) get the understanding of the product
2) make the changes
3) get the changes right.
OO seems to help in all these areas.
===

In reality the thing I find is that the code I write is
really a hybrid of procedural and OO where the more procedural
stuff is usually part of the objects that make sense.

I think after experience, people should learn to avoid OO for
OO sake.
Maybe I'm not a purist. That's fine. I'm ok with it.

What I find for reality is that in a hybrid model, you
use OO for the things that make sense and use procedural
programming for some things.

My goal's for almost everything I do follows that old
80/20 rule where you stop doing something when you
start getting diminishing returns.

"20% more design time will save 80% time during the updates"
That 20% of the design time should be focused on complex,
frequently updated, problematic or areas suited to OO.
(Suited to OO is something that comes with practice).

i.e. spend 420 work units in order to get updates
that cost 33 work units. ( my update numbers 65 - ((65-25)*.8)

Reality of these numbers are certainly arguable too.


"The stuff we call "software" is not like anything that human society
  is used to thinking about. Software is something like a machine, and
  something like mathematics, and something like language, and
  something like thought, and art, and information...
  but software is not in fact any of those other things."
Bruce Sterling - The Hacker Crackdown

Fred A. Kulack  -  AS/400e  Java and Java DB2 access, Jdbc, JTA, etc...
IBM in Rochester, MN  (Phone: 507.253.5982   T/L 553-5982)
mailto:kulack@us.ibm.com   Personal: mailto:kulack@bresnanlink.net
AOL Instant Messenger: Home:FKulack  Work:FKulackWrk

+---
| This is the JAVA/400 Mailing List!
| To submit a new message, send your mail to JAVA400-L@midrange.com.
| To subscribe to this list send email to JAVA400-L-SUB@midrange.com.
| To unsubscribe from this list send email to JAVA400-L-UNSUB@midrange.com.
| Questions should be directed to the list owner: joe@zappie.net
+---