As I was saying to someone at the EuroPython conference; I had been struggling as a Python programmer to get beyond the map, filter (and reduce) functions. Functional programming "done", I just didn't know where to go next to apply more learning having reached this glass ceiling.

After three days of overwhelming talks, the list is now ballooning with new areas for me to approach.

Russel Winder's talk on Gil was a warning about difficult methods to parallelise code and recommended the multiprocessing module.

From a technologies perspective, Esteve Fernandez's talk Twisted, AMQP and Thrift was the most provoking to give-a-try in some free time. I've always felt daunted by the complexity of Java RMI tools, Twisted, AMQP and Thrift would surely struggle to be more complex than their Java cousins.

Rob Collins had a talk on his FilterPype framework which seemed like a fantastic development of the piping mechanism in unix. It was this talk which started the "I remember something about that from uni" klaxons going off and Bruce Eckel continued in this, during in his presentations.

The main theoretical subjects which I heard mentioned again and again, but feel ashamed to know little about were:

  • Continuations
  • Subroutines
  • Closures
  • Generator expressions

I also heard the following Python tools / techniques mentioned

and should investigate these

  • wsgi
  • zope
  • @memoize decorator to do as haskell does when first calculating factorial of 6 and then 7 after (factorial(7) does 7 * _cached_factorial(6))
  • (Also Lua, a different language)

David Jones' What Sucks About Python was another great talk on positive (and some negative) features of python where he demonstrated continuation issues which I need to follow on my own interpreter.

(I also heard Kamaelia sporadically for concurrency)

A thoroughly worthwhile use of three days. I am sure the event will help me develop as much as a programmer as a pythonista