上手です。
XPengの Robert C. Martin の型チェックについての発言です。ご参考まで。
＃oosquare ネタですね
＃言語は弱いので、訳はおかしいです。悪しからず。
Date: Thu, 17 Aug 2000 21:55:28 -0500
To be clear about this, I have walked both the checked and unchecked
side of
the street for a long long time. I learned Smalltalk, Logo, and Prolog in
the early '80s; and did a lot of playing around with them. I did a *lot* of
C before that (not type checked, but still statically typed!!!) And I
worked heavily in C++ and later Java.
型チェックの両方で長いこと、いろいろやってきた。特にC++と最近はJava。
Until recently I was of the opinion that statically typed languages were
essential for any industrial software effort simply because they found so
many errors. I had written C for too long to be unafraid of type errors.
最近まで、静的型付け言語は産業用のソフトゥエアには基本だと思っていた。Cで
余りにも長い間、型エラーを怖れずに書いていたし。
What has changed my mind about this is the near fanatic dependence of XP
upon unit tests. The unit tests take the place of, and far surpass, the
benefits of static type checking. This make the benefits of dynamic typing
far exceed the costs.
この考えを変えたのはXPの狂信に近い単体テストへの依存だ。単体テストは静的
型チェックの恩恵に替わり、さらにはるかに上回る。
So, IMHO, static languages are heading into a slow decline. C++, Java, and
even (sigh) C# will continue strong for many years to come. But I expect
that the next wave of new languages that capture the attention of the
industry will be dynamically typed. Python may be the first of many to
come. We may even see a resurgence of Smalltalk.
静的言語は少しずつ地盤沈下の方向だ。C++、Java、そしてC#もずっと強いだろう。
しかし、産業の興味をとらえる新しい言語は動的に型付けされているだろう。
Pythonが最初になるかもしれない。（あの）Smalltalkでさえ復活するかもしれな
い。
Robert C. Martin