He isn't different in person if you get him going. In fact, he'll happily admit "he likes to poke people"

Basically, he's a very very very successful troll :)

> "I don't doubt Linus would be less brutal in person."

My personal rule of thumb is do not say something written that I wouldn't say in person, especially if I am using a profanity language.

That may be your rule, but it is definitely not the norm. I've known countless people, including the infamous Erik Naggum, that are far more abrasive via email than they are in person. Erik was, in person, a sweet and thoughtful person that would never attack someone. On email or usenet, he was on the order of Linus at attacking people.

So, while it's your good rule of thumb, my observation over the years is it is not followed by many.

I saw lots of email from Erik whose viciousness went far above and beyond anything I've ever seen from Linus. And Erik was extremely unsuccessful at organizing collaborative efforts; Lisp languished for over a decade under his dominance. I don't think it's a coincidence that Erik's retiring from comp.lang.lisp was roughly concurrent with the revival of the Lisp family of languages.

First, Erik wasn't Lisp. He didn't even work for a Lisp company, except for consulting for Franz briefly. Second, your cause and effect meter is way off. The resurgence of Lisp happened because there were a LOT of people out of work in 2001 and many of them started working on CMUCL and SBCL.

Erik's influence ended outside of comp.lang.lisp and the influence of that usenet group was fairly small, in the greater scheme of things.

Lisp didn't resurge because SBCL started being developed again. SBCL, although I like it, is still relatively little used, in the overall scheme of things, a scheme which includes Clojure, elisp, Racket, Clisp, Chicken, TinyScheme, Guile, Arc, MCL, and so on. Lisp resurged because new hackers started using Lisp again, maybe in part because of pg, but I think in large part because they didn't get driven away by Erik. Lisp companies are comparatively irrelevant.

comp.lang.lisp was very important prior to the founding of Stack Overflow, EmacsWiki, and so on. It was a lot more important to Lisp than any Lisp company, and maybe more important than all Lisp companies put together.

comp.lang.lisp has long lost any importance. Most Common Lisp users have moved on. The communication infrastructure shifted to the mailing lists ( http://dir.gmane.org/index.php?prefix=gmane.lisp ) and the #lisp IRC channel. That's where the real work was done. Even much random talk now has been moved since quite some time to reddit / lisp, stackoverflow, planet lisp and even Twitter #lisp.

comp.lang.lisp is long dead.

Indeed, comp.lang.lisp is now irrelevant; but it wasn't irrelevant at the time Erik retired from it, despite his strenuous efforts.

However, I wasn't talking about only Common Lisp users, but about users of all Lisps. Elisp and Clojure have similarly-sized Freenode channels #emacs and #clojure, each substantially bigger than #lisp. Various estimates of the number of Emacs users put them around 5–50k people: there are perhaps around 15000 Elisp users http://stackoverflow.com/questions/3311227/emacs-user-base-s... or 0.5% of Ubuntu users using Elisp http://programmers.stackexchange.com/questions/35312/emacs-u... (some tens of thousands) or 5300 Github users with Elisp repositories http://www.quora.com/How-many-people-use-emacs-per-day?share....

How is using an editor making them Lisp users? Are users of Google's flight search engine, written in Common Lisp, making them Common Lisp users? Are AutoCAD users Lisp users?

How many Emacs users just use it? How many already need help to configure .emacs ?

Typically AutoCAD users are not AutoLISP users. (By the way, AutoLISP conses are also immutable, although you have complained that the immutability and resulting "persistence" of Clojure lists counts against calling it a Lisp.) But typically Emacs users are indeed Elisp users, because they read, write, and modify Elisp code. That's why there are several thousand Elisp repositories on GitHub.

     My personal rule of thumb is do not say something written that I wouldn't say in person, especially if I am using a profanity language.

Obviously not Linus' rule... Anyone who's ever followed the kernel dev mailing list knows this.