| Imagine it's, like, 1980 - or even earlier - and you can work in a language roughly as nice as Clojure, except the rest of the world is stuck working with pre-ANSI C or Pascal or FORTRAN or COBOL or raw assembly language. There's no Python or Java or C# or Ruby or Perl or Haskell or Scala or Kotlin or Rust or JS/TS. Nothing really resembling our modern idea of a high-level language. (OK, there was Smalltalk. Let's ignore Smalltalk. Lord knows everyone else did.) That'll alter your perception of reality a bit. Here they were, in possession of a tool massively more powerful - and more elegant - than what everyone else is using. And moreover, everyone else took a look at it and turned their noses up. Clearly, you and your fellow Lisp programmers are a different breed, capable of seeing further than the rest of the unwashed masses. In a word, you were better than them. It sounds like I'm being disparaging, but to a certain extent, I don't even think this was totally a wrong attitude to have. Elitist, definitely, but not wholly unwarranted. Lisp really was - in terms of expressiveness, anyway - really that far ahead of the competition. And yet somehow that competition won. The world is cruel and unjust. So Lisp becomes a kind of Us v. Them cult: if you've heard the good word of McCarthy, you're one of Us. If not, you're best ignored - too stupid to possibly have anything worthwhile to say. (If you think I'm exaggerating, spend some time reading the words of Usenet Lisp institution Erik Naggum - R.I.P. - who serves as the most extreme but hardly the only example.) This blinded Lisp diehards to the outside world, which slowly but surely, in many respects, began to catch up or even exceed Lisp. The other thing is - not only is Lisp a powerful language, at its core is a beautiful and simple and expressive mathematical idea. Combine that with the way macros allow you to extend the language virtually infinitely, there can be a near religiosity at the heart of Lisp - from one lambda all things depend. Lisp isn't just good engineering - it's a glimpse at the fundamental nature of computation, of the universe itself. I'm not going to sit here and tell you that this is somehow a terrible thing, per se. But it can be incredibly alluring to the right kind of mind, and once you're in its thralls it's hard to get out. You might be working with the tool, but in another sense the tool is working with you. A Siren Song. |
Early Internet discourse around programming was dominated by people who had ties to elite universities in the 1980s, who yearned for the times when the US Government was throwing an abundance of money to the AI industry of the time.
They were the ones rubbing elbows with researchers from MIT, Stanford, Harvard, and Berkeley, who were using specialized hardware and software beyond the capabilities available to that of developers working on more mundane applications, all graciously funded by DARPA initiatives.
That experience was, in truth, unrelatable to young people reading the recollections of ESR and RMS of the period, the in-jokes of these people, their ideas and interactions, but the tales of Lisp, the Lisp hackers and their fabled Lisp machines would be extremely appealing to someone who was very passionate about programming, striving for excellence as a programmer, and to advance in life through merit. Paul Graham would seal the deal with his essays.