Prev: OOP is own private Cleveland (Re: OO Books for newbies?)
Next: SOA vs OO
From: topmind on 14 May 2007 13:20

Daniel T. wrote:
> topmind <topmind(a)technologist.com> wrote:
> > Daniel T. wrote:
> > > topmind <topmind(a)technologist.com> wrote:
> > > > "Daniel T." <danie...(a)earthlink.net> wrote:
> > > > > topmind <topm...(a)technologist.com> wrote:
> > > > > > Daniel T. wrote:
> > >
> > > > > > > I think it is important to remember why Simula was
> > > > > > > developed in the first place. If your problem domain does
> > > > > > > not involve doing simulations, then OO may very well not
> > > > > > > be appropriate.
> > > > >
> > > > > > If the industry agreed that OO generally only shines for
> > > > > > physical simulations and not push it into nearly all
> > > > > > domains, then I would not be ranting against it.
> > > > >
> > > > > Wait, why the restriction of "physical" simulations? Simula
> > > > > and other OO languages are excellent at modeling simulations,
> > > > > "physical" or otherwise.
> > > >
> > > > I am not sure what your point is. There appears to be context
> > > > confusion.
> > >
> > > Simulation: the imitative representation of the functioning of one
> > > system or process by means of the functioning of another. For
> > > example when a computer program simulates the processes employed
> > > by a book keeper, bank teller, or stock trader...
> >
> > Yes, but in practice we don't do it that way. The manual processes
> > are designed to reduce human errors, not necessarily geared toward
> > efficiency. One often changes the way the process is done when
> > computerizing it.
>
> It is still a simulation. Simulations are not expected or required to
> perform the process the exact same way, as long as the results are
> equivalent.

If we have N different models that *all* produce the right answer
given the same set of input, how do we evaluate which is "better"?
(Let's ignore processing speed for now.)

(I replied to this earlier, but it seems to be vacationing in the
Usenet Black Hole)

-T-

From: Daniel T. on 14 May 2007 19:32
topmind <topmind(a)technologist.com> wrote:

> If we have N different models that *all* produce the right answer
> given the same set of input, how do we evaluate which is "better"?

The answer, of course, is highly dependent on context. How many people
are working on the project? What is their general skill level? Is the
project projected for continual development or will there be a drop-dead
date? I'm sure there are many more variables...
From: topmind on 14 May 2007 20:31
On May 14, 4:32 pm, "Daniel T." <danie...(a)earthlink.net> wrote:
> topmind <topm...(a)technologist.com> wrote:
> > If we have N different models that *all* produce the right answer
> > given the same set of input, how do we evaluate which is "better"?
>
> The answer, of course, is highly dependent on context. How many people
> are working on the project? What is their general skill level? Is the
> project projected for continual development or will there be a drop-dead
> date? I'm sure there are many more variables...

Is there too many variables, too many *subjective* variables, or both?
If there is a *single* significant subjective variable, then we will
never turn paradigm comparison into a science even IF we can measure
the other variables.

Put another way, can the psychology of developers ever be removed from
a good comparison?

-T-

From: Daniel T. on 16 May 2007 07:05
topmind <topmind(a)technologist.com> wrote:
> On May 14, 4:32 pm, "Daniel T." <danie...(a)earthlink.net> wrote:
> > topmind <topm...(a)technologist.com> wrote:
>
> > > If we have N different models that *all* produce the right answer
> > > given the same set of input, how do we evaluate which is "better"?
> >
> > The answer, of course, is highly dependent on context. How many people
> > are working on the project? What is their general skill level? Is the
> > project projected for continual development or will there be a drop-dead
> > date? I'm sure there are many more variables...
>
> Is there too many variables, too many *subjective* variables, or both?

I don't know that there are too many variables and I don't think that
any of the variables are subjective. So my answer to the above is
"neither".

However, the variables are still important. Paradigms cannot be compared
in a vacuum. You yourself have admitted as much; when I first started
reading your posts here back about 10 years ago (has it been 10?) you
said "OO bad", now you have to put a whole bunch of qualifiers on in
what contexts you are considering when you say that.
From: topmind on 16 May 2007 16:22

Daniel T. wrote:
> topmind <topmind(a)technologist.com> wrote:
> > On May 14, 4:32 pm, "Daniel T." <danie...(a)earthlink.net> wrote:
> > > topmind <topm...(a)technologist.com> wrote:
> >
> > > > If we have N different models that *all* produce the right answer
> > > > given the same set of input, how do we evaluate which is "better"?
> > >
> > > The answer, of course, is highly dependent on context. How many people
> > > are working on the project? What is their general skill level? Is the
> > > project projected for continual development or will there be a drop-dead
> > > date? I'm sure there are many more variables...
> >
> > Is there too many variables, too many *subjective* variables, or both?
>
> I don't know that there are too many variables and I don't think that
> any of the variables are subjective. So my answer to the above is
> "neither".
>
> However, the variables are still important. Paradigms cannot be compared
> in a vacuum. You yourself have admitted as much; when I first started
> reading your posts here back about 10 years ago (has it been 10?) you
> said "OO bad", now you have to put a whole bunch of qualifiers on in
> what contexts you are considering when you say that.

It's called "learning". I started off with 2 primary assumptions that
I slowly found to be wrong, or at least non-reliable. In particular
I've learned:

1. Psychology is key: Every person's mental model of the world and how
they view and deal with code, grids, languages, etc. is vastly
different from each other.

2. What works well for the domain I am familiar with may not be the
case in other domains.

I think software engineering is *inharently* tied to human psychology
on both a global and local (individual) level. You appear to imply[1]
all that it takes is an expert "judge" to render a professional
opinion, and that that is as good as or nearly the same as objective
metrics. It is not. If one cannot dissect out the specifics,
something is probably wrong or subjective about the technique.

[1] Without details, it is difficult to better characterize your
method of measurment.

-T-

 | 
Pages: 1
Prev: OOP is own private Cleveland (Re: OO Books for newbies?)
Next: SOA vs OO