Touche. :)

I think you're right. I already found a task for Python to do, spent a couple weeks (off and on) refining it, and at the end I seemed to have used all of the prominent features of Python somewhere in that script. At the end of that, I had this vague feeling that I now "knew" Python and wasn't sure what was left to learn.

So, you are confirming my vague feeling and I can now move on to Scala and decide whether or not to switch to Arc once it emerges. :)

Since you're feeling cocky, have a look at the Zope/Plone source. You'll almost certainly realize you don't know Python (and possibly that you don't actually want to).

There is a difference between a language and its common idioms. "Generator expressions" is a good phrase to test how much people know. For example: http://norvig.com/sudoku.html

Python can definitely fit in your head, as far as the language. Reading lots of code is a good way to learn the idioms.

I wasn't actually referring to the usual Python idioms, but if you're unfamiliar with Zope/Plone, you can be forgiven for not knowing that. The Zope developers (and to a lesser degree Plone) don't actually seem to like Python, and so they've reinvented many common language features and libraries. Partly this is historical...the Python that Zope was originally written in didn't have most of those features or the standard library that exists today. It's one of the problems of developing a large codebase in a language that is still in rapid flux and doesn't mind throwing out backward compatibility (my Perl code written before Zope existed still runs on modern Perl, while practically none of my Python code from 1.5 through 2.2 runs today unmodified).

But, regardless of the very sound reasons for it to be the way it is, it's a very intimidating codebase, and a really good case study in the "reinvent everything" mindset. (It's quite powerful, and I've worked with several of the guys who built it, as I did contract work for Zope Corp for a few years, and I have the utmost respect for their talent and intellect. They're some of the smartest developers I've ever met. But it's not quite like anything else I've ever seen when it comes to having a learning cliff, even with some pre-existing Python knowledge.)

Yah, I assumed they were doing something fancy, not knowing what they actually are doing.

I'm told a great place to read good code, both Python and C, is to read the code that makes Python.

Yes, C is another good example. The language is fairly small, and it's not terribly difficult to learn the language itself and write your own programs. But to be able to look at an arbitrary piece of C code and understand all of the implications is another matter entirely.

This is true of any language to some degree, although some are worse than others, and if the code was purposefully written in some non-traditional way it can be even harder.

It is most certainly not true about C++.

Touche'; I should have grouped my sentences differently. C++ is not a small, easy-to-learn language.

I'm reminded of Duff's Device http://en.wikipedia.org/wiki/Duff%27s_device

  send(to, from, count)
  register short *to, *from;
  register count;
  {
    register n = (count + 7) / 8;
    switch (count % 8) {
    case 0: do { *to = *from++;
    case 7:      *to = *from++;
    case 6:      *to = *from++;
    case 5:      *to = *from++;
    case 4:      *to = *from++;
    case 3:      *to = *from++;
    case 2:      *to = *from++;
    case 1:      *to = *from++;
               } while (--n > 0);
    }
  }

I knew about list comprehensions, but generator expressions add some new tricks. And I also now have a clearer idea of what constraint propagation means.

Thanks for the link!