Lisp is for Entrepreneurs (opens in new tab)

Are you saying that if someone built a house with an inflatable mallet and someone else built one with a nail gun there would be no difference of quality in the resulting house, or that one wouldn't be completed sooner, or be more extensible? (Yes, you can laugh at me.)

I agree with you that all languages are tools (I love working with all kinds, actually). But aren't some tools better at some things than others? Or what would we need so many for?

I'm not saying Lisp is the nail gun and everything else is an inflatable mallet, but I do think there is a difference.

What I got out of the article is that Lisp is a tool that worked pretty darn well in a couple cases. I feel that's a good thing, because I kind of like Lisp.

Languages are not just hammers. They are universes for thinking in.

So you're using Rails... What if I told you all your competitors were using Java and Spring? Gives you a surge of confidence, doesn't it? That's the feeling Lispers have about Rails programmers, for the same reason.

I have personally seen Lisp stamped out for political reasons. I had been tasked with generating C code from a textual description of a finite state machine. With my boss's permission (also a Lisp programmer) I used scheme to take in the descriptions and output C code. It hooked right into the Makefiles and worked well. Despite the fact that the solution worked and that there were several engineers on staff who were Lisp programmers this did not make it through review. A senior engineer who had been with the company for a long time (who'd never programmed in Lisp) had major objections. He pulled the right strings and we were forced to reimplement it all in C. I suspect one of real problems for Lisp is the presence of this sort of anti-Lisp person in the industry.

Yes, it's an education barrier. BRL (Scheme adapted for web apps) has survived numerous attempts to quash it at my employer over the past 7 years, as different managers have needed educating.

The other education barrier is with programmers. The "Wow, this is simple and useful!" response comes in 15 minutes using Python or Ruby. With Scheme or CL it comes in a week or two, though it's stronger. I don't know if s-expr notation will ever have immediate appeal, or if it needs to grow on everybody.

IIRC their main issues were threading instability in CMUCL on BSD or mac, which I believe has since been fixed in SBCL. In general, SBCL has come a long way, partly due to companies that use CL (e.g. ITA) funding its development. CL is in many ways much more production-ready than it used to be for that reason. It still has rough spots, but my experience is vastly improved over what it was 1.5 years ago.