Jonah Thomas wrote:
> Sure. Your advice is good but it doesn't fit my needs. Let's try a
> similar discussion in a different context. I don't like quantum
> mechanics because by all accounts it does not make sense. People can
> solve problems with it but they get no sense of what's possible from it,
> and experienced QM users tend to have no useful intuition about problems
> they have not yet formally solved. In some contexts it looks like a dead
> end. And by some accounts it is sort of invalid, some of the results
> come from things like fudge factors and taking the first few terms of
> series that aren't sure to converge and so on. I don't like it but I
> don't have much to do with it. If I were to go onto a physics newsgroup
> and say I don't like it because it doesn't make sense and I want
> something that will get similar answers that does make sense, I expect
> I'd get a response that's very similar to yours. "Before you criticise
> QM, you need to learn all about QM. Learn all the math. Learn how to
> solve QM problems, and solve a lot of significant problems, not just
> simple textbook examples. Only then will you be qualified to discuss the
> issues with it." But I wouldn't want to do that. Probably I have only
> around 30 productive years left. Should I spend, say, 15 of them
> learning all about something that I have every reason to think will not
> make sense after I understand it? No. If I wanted to go that route I'd
> do better to look at the newer alternatives first. Start with things
> that make some sort of sense, and look for ways to improve them so they
> fit the data.
Like Albert said before about FORTRAN and people learning only one
language:
If something is difficult to learn and takes a long time until you get
there, it will make you want to stick with it. After all, it took so long
to get there, so whereever else you go, it probably takes just as long,
and
you don't know if you'll be successful. You also started to learn it when
you were young and your brain was agile, and now, after you mastered it,
you are old, and learning new things are more difficult.
Solutions to the inherent problems of this domain will *not* come from
insiders. QM experts will not provide a replacement theory that makes
sense - it is highly unlikely that they will find value in such a theory
if
they see it first. Planck said more than a hundred years ago, that new
theories often require the professors to be replaced before they are
accepted.
OOP had been announced as "silver bullet" for the software problem. As
Fred
Brooks states in his article "no silver bullet" (20 years after first
publication of "The Mythical Man-Month"), it isn't a silver bullet. Some
problem domains map quite well into the OOP way of solving problems, like
GUIs. But the best way to deal with complexity is still "keep it simple".
Then, as fallback solution, you can deal with the unavoidable complexity
by
using OOP - if the problem fits into this solution space.
--
Bernd Paysan
"If you want it done right, you have to do it yourself"
http://www.jwdt.com/~paysan/
|
