It’s in AutoLisp (within AutoCAD). For the same reason.

Which actually fits your ideas. The payoff from Elisp is so much higher than from the other two that it hardly seems worth it to mess with them. Or anything else if you hope to edit and move around data at the end of the day. Web pages are obviously a different story.

Sounds like an oxymoron to me.

“closures, active records, lambdas, embedded SQL a la LINQ, etc. etc. and so those are the kinds of features we put into Wasabi…”

Sure, Joel, and all starting from VBScript. Give me a break!

So is FogBugz profitable or is it a vanity shoppe financed by stock options Joel sold when he left Microsoft? Why hasn’t FogBugz gone public?

Despite its wandering around in a number of different directions, this is a fine essay. You touched on a number of points relevant to your main topic: the use and misuse of niche programming languages, and the choice to use or not use them.

Mayson

I think that markus’s first point has some merit. Some languages simply are better than others. There are a lot of posts out there making statements such as “LISP macros ultimately make programs unmaintainable”, but these are mostly rationalizations by authors who have never used any of the advanced languages (and who are rightly annoyed by the insinuation that they cretins for not using such languages).

So why don’t we see more good stuff from Haskell and LISP lovers? Most of us are programmers by trade. Even those amongst us who are passionate about programming are pretty pooped at the end of a programming day. That doesn’t mean that we don’t play with our favourite languages; only that it’s hard to start on big projects.

So what about the BASIC programmers? Don’t they have day jobs (that is, aren’t many of them students)? I don’t know. But their goals are quite different. I don’t have a desire to write 2D side-scrollers (for example). In a 3D game, I’d be more interested in the 3D engine than the game. Or better yet, in the AI.

We don’t see too many BASIC games with advanced AI. That’s not an indictment or a condescending remark. But clearly the goal of a BASIC programmer is to produce a working (fun, cool) game, whereas a language “snob” might be interested in other areas that are not so amenable to the quick feedback cycle.

If LISP was more widely used, one would see a heap of cool stuff written in LISP. In terms of deployment and libraries, we have languages/implementations such as CAL, Clojure, SISC and Scala on the JVM alone. I’ve shown friends and colleagues these languages; each of them rationalized the non-use of these languages away in a different (and unconvincing) way. Also, don’t forget what a negative a formidable reputation can have on a language (“Person A: I want to write my game in Scheme. Person B: WHAT!? Do you know how hard that language is to use? Better stick to BASIC.”)