Yesterday I did a piece of programming that was a complete, utter and absolute waste of time.

It was almost completely useless; it was a vast expenditure of time to achieve something automatically which (a) wasn't that important anyway, and (b) I could have achieved by hand in a fraction of the time it took me to develop a general method for automating it. I tried enthusing about it to a couple of people last night and they told me I was completely mad.

Perhaps I am, I suppose. But I did it largely because it had recently occurred to me that I've become more concerned with ends than means over the years I've been an obsessive geek. I used to write code with no thought of how useful it would be to whom, but simply because it was fun; because the actual act of designing, writing and debugging a complex piece of code was a reward in itself, and because the feeling when a creation of mine finally stuttered into life and started giving sensible answers was a little like what it must feel like to be able to do magic.

These days it seems to me that it's almost all about the end result. I write programs because there are problems I want solved, and the satisfaction comes after the coding, when I sit back and bask in the pleasure of not having to deal with whatever the problem was any more. And I write other programs - such as PuTTY - because they're popular, and because it makes me feel I'm doing something useful for the world, and because I like the idea that lots of real people out there have their lives made that little bit more pleasant by what I spend a hefty chunk of my spare time slaving away over. But I didn't start out that way, at all; I started out because the act of programming itself was enjoyable. If I'd gone into coding because I saw all sorts of problems that needed to be solved and realised that learning how to solve them was the only way to make them go away, it would have been a pretty joyless discipline for me to have dedicated my life to.

A whole chunk of yesterday evening was completely wasted, by any objective measure. But the particular piece of coding I did was particularly enjoyable in total disproportion to its usefulness, and so I think that subjectively, it was the most worthwhile thing I've done in months: it reminded me of the other way it's possible to enjoy programming, and that just might be more valuable to me in the long run than any number of solutions to niggling little practical problems.