Programming languages in economics: Cool research, bro, but what about, er, economics?

Machines that go ping

Let's apply this same idea to other areas of coding life. If you're Google or Facebook, with petabytes of data running through expensive data centres every minute of the day, then you might well spend a certain amount of time really optimising your code to make sure that running time was something that you'd optimised.

Remember the DOS game Another World? Your correspondent helped port it to Windows

If you were trying to make a machine that goes “Ping!” to alert the CEO's secretary that he's got a new email, then any old bodge job will do, even Java possibly.

Both your knowledge of game and standard economic theory are leading us to the same critique of this paper. You don't want to look at just one part of the process or system to work out where it's worth expending more effort to gain a faster or better answer. You want to look at all of the resources in use and then optimise the use of the scarce resource or resources. That's – in both senses – the economic way.

Barrow boy that I am, I once started a little business on this very thought. Back when the Berlin Wall came down (I've written about this for El Reg here before), I mused on the following point. We've got 386 computers and some other users over there are still stuck with 8086s. We've got computer games and they've got computer games. Games, meanwhile, are famous resource hogs. So, are they compensating for shite hardware by concentrating on the scarce resource? After a bit of travelling around and talking to people I found out that I was correct. (I used to know the Russian, Polish and even Hungarian for “Do you know a programmer?” - that phrase and “Have a beer” in the correct language worked surprisingly well as a research method.) Eastern games programmers were writing much tighter, more ordered – heck just better designed – code than their developed world counterparts.

Much of this was simply because they had to to be able to squeeze any game play at all out of the machinery that their customers had available. They were concentrating on the scarce resource and optimising for that.

The business idea was simple and worked well until games budgets started to balloon and no one wanted to let development out of their sight. We'd take Western code and go optimise it for more difficult or more limited running environments – get Amiga games to run on C64s, that sort of thing. The only one I'm actually credited on is the Windows port of Eric Chahi’s seminal game work Another World, as the producer. That's one that everyone said could not actually be done but it was possible with people who already concentrated on optimising that scarce resource.

But back to the basic point. You don't choose a programming language on runtime alone. You look at the whole process and then optimise for whatever the scarce resource is. It's just most odd that economists, who are supposed to concentrate on the idea of allocating scarce resources, don't do the same in an economics paper about programming economics. ®