Bob La Quey wrote:
I suggest we ask people who speak English as a
second or third language what they think about
the issue. Are any of these other languages worth
more than two weeks effort?
There is a difference. I can *see* what learning a second language gets
me. I can talk to people; I can read literature I didn't have access
to, etc. And, even so, sometimes keeping the motivation going for
learning another language gets hard.
This visibility is *not* true for computer languages.
Is it really the case that if you cannot learn a
computer language in two weeks then there must be
something wrong with the langauge. Or is this a
statement about how intellectually lazy most
programmers are?
Yes and yes. ;)
In my opinion, a better analogy is the use of a chisel or a plane.
I can use a plane to solve my problem fairly quickly if I have to and it
won't give me much grief. It won't be pretty, but it will work.
If I want pretty or productive, I need to learn to use the tool. That
takes time. In addition, my quality and productiveness grows in some
proportion to how much time I spend to learn to use the tool.
What the author was objecting to is the fact that the language is
actively impeding him from doing simple things at the beginning. That's
*not* the sign of a good tool.
Say what you will about the Lisp/Schemes, but if you *want* to write
iterative code, you can. If you want lots of side effects and
assignements, that's okay, too. If you want clean code and closures,
okay, we'll work with you there, as well.
Lisp started to get over its "my way or the highway" attitude and, lo,
and behold, people started using it. Go figure.
And, in fact, even a tool like a plane shows the same pattern. A
Stanley Combination Plane is a work of art that enables a good craftsman
to cut *lots* of different things with a single tool. Yet, it requires
a *lot* of time and effort to achieve that level of skill. As such, a
few people rave about them and the vast majority curse at them. With
the benefit of time in hindsight, it seems that the majority were right.
The tool was too complicated and there were better ways to do the same
tasks.
So it goes.
-a
--
[email protected]
http://www.kernel-panic.org/cgi-bin/mailman/listinfo/kplug-lpsg
