TITLE: Learning and New Kinds of Problems
AUTHOR: Eugene Wallingford
DATE: October 13, 2011  3:10 PM
DESC: 
-----
BODY:
I recently passed this
<A HREF="http://weblog.raganwald.com/2007/02/but-y-would-i-want-to-do-thing-like.html">
   classic by Reg Braithwaite</A>
to a grad student who is reading in the areas of functional
programming and Ruby.  I love how Braithwaite prefaces the
technical content of the entry with an exhortation to
learners:

<BLOCKQUOTE><EM>
... to obtain the deepest benefit from learning a new language,
you must learn to think in the new language, not just learn to
translate your favourite programming language syntax and idioms
into it.
</EM></BLOCKQUOTE>

The more different the thing you are learning from what you
already know, the more important this advice.  You are already
good at solving the problems your current languages solve well!

<BLOCKQUOTE><EM>
And worse, when a new tool is applied to a problem you think
you know well, you will probably dismiss the things the new
tool does well.  Look at how many people dismiss brevity of
code.  Note that all of the people ignore the statistics about
the constant ratio between bugs and lines of code use verbose
languages.  Look at how many people dismiss continuation-based
servers as a design approach.  Note that all of them use
programming languages bereft of control flow abstractions.
</EM></BLOCKQUOTE>

<IMG SRC="http://www.cs.uni.edu/~wallingf/blog-images/computing/y-combinator.jpg"
     ALT="Real programmers know Y."
     WIDTH="200"
     ALIGN="right" />

This is great advice for people trying to learn functional
programming, which is all the rage these days.  Many people
come to a language like Scheme, find it lacking for the
problems they have been solving in Python, C, and Java, and
assume something is wrong with Scheme, or with functional
programming more generally.  It's easy to forget that the
languages you know and the problems you solve are usually
connected in a variety of ways, not the least of which for
university students is that we teach them to solve problems
most easily solved by the languages we teach them!

If you keep working on the problems your current language
solves well, then you miss out on the strengths of something
different.  You need to stretch not only your skill set but
also your <EM>imagination</EM>.

If you buy this argument, schedule some time to work through
Braithwaite's derivation of the
<A HREF="http://en.wikipedia.org/wiki/Y_Combinator">
   Y combinator</A>
in Ruby.  It will, as my daughter likes to say, make your
brain hurt.  That's a good thing.  Just like with physical
exercise, sometimes we need to stretch our minds, and make
them hurt a bit, on the way to making them stronger.
-----