Consider the case of Scheme, again. Since making Scheme object-oriented is so easy, many Scheme hackers have done so. More to the point, many individual Scheme

Consider the case of Scheme, again. Since making Scheme object-oriented is so easy, many Scheme hackers have done so. More to the point, many individual Scheme hackers have done so. In the 1990s, this led to a veritable warehouse inventory list of object-oriented packages for the language. It's funny that something quite similar can be said about today's arguably most popular and ubiquitous language: JavaScript. Each of MooTools, ExtJS, Prototype etc. sport their own model of OOP, with minor or even major differences. Yet it doesn't seem to be so big of an issue that it prevents cool stuff from being done in JS. For one, I don't think many people would claim that you need some implementation of OOP (i.e. classes, inheritance 'n stuff) to effectively program in JS. The readily available mechanisms are perfectly sufficient to just start coding what you want to code and doesn't require "amending" the language first. JavaScript code generally seem to be fine with that and happily proceed to create another batch of coolstuff.js. But according to the article, Lisp hackers have totally different mindset: they first "create a language" and only then proceed to use it to actually code stuff. (There is quote about giving man a Lisp...) If that's indeed true, I don't think it's really an issue of the Lisp itself, i.e. it's all-too-great power. After all you can just proceed into writing your program rather than writing a language first - like in JS. If Lisp coders fails to do that, I would say it's a problem of Lisp coders.