I'm sure I'm not the only guy who started learning Lisp after reading Paul Graham's essays, and I'm going to keep using Lisp for several reasons, but the momentum just isn't there.
"Robert and I both knew Lisp well, and we couldn't see any reason not to trust our instincts and go with Lisp. We knew that everyone else was writing their software in C++ or Perl. But we also knew that that didn't mean anything. If you chose technology that way, you'd be running Windows. When you choose technology, you have to ignore what other people are doing, and consider only what will work the best."
"This is especially true in a startup. In a big company, you can do what all the other big companies are doing. But a startup can't do what all the other startups do. I don't think a lot of people realize this, even in startups."
[...]
"If other companies didn't want to use Lisp, so much the better. It might give us a technological edge, and we needed all the help we could get."
Seems like a lot of startups do do what the others do tech wise and use Ruby which seems to work out ok.
I have optimism for lisp.
EDIT: always in SBCL and CCL
It is one of the most active implementation of Common Lisp (other being Clozure CL) and one of the fastest thanks to its support of optional type hinting (which it had long, long before gradual typing become in vogue recently). With some type hinting, it is possible to generate a code that's almost as fast as C (in certain cases, of course, but still much better than other implementations).
>Nobody seems to have done any serious work in SBCL in over a decade.
Could you please elaborate on what you mean by this?
SBCL releases a new version that contains both enhancements and bugfixes every couple of months, with the latest version (version 1.4.9, released on June 28, 2018, ~5 days ago) came one just one month after the previous version, 1.4.8, so I really don't understand what you mean when you said that no one "seems to have done any serious work in SBCL in over a decade".
You can also see all previous releases and their changlog here: http://www.sbcl.org/all-news.html.
Or did you meant to say that no one use SBCL to do serious work? Well, with some googling, I'm sure you'll be able to find that, while not many, there are companies that use SBCL (and other Common Lisp implementations) in production and other non-trivial works (not to mention various freelancers who use SBCL to put foods on the table). Some of those companies are also involves with the development of SBCL as well.
>Basic things are either completely missing or broken.
Care to list them? One of the strength of SBCL is that the devs are very responsive, especially to bug reports, so I'm sure they would be more than happy to fix the problems that you encountered.
When is the last time you use SBCL, or Common Lisp in general? Quicklisp (https://www.quicklisp.org/beta/), which is a modern package manager that supports resolving dependency and works across almost all currently active Common Lisp implementations (and not just SBCL), has been a thing for years now.
>The basic HTTP server, hunchentoot IIRC, crashes after it serves its first request
Would love to see the backtrace to see the reasons (and to submit a bug report if the issue warrant one), as from my experience Hunchentoot is very stable. I regularly used it in my freelance jobs without any issues.
What are those things? Name them.
The commodification of programming and proliferation of languages like Javascript and Python means not only that the "masses" lack fundamental understanding of Computer Science but that they have been conditioned to go after superficial ease-of-use rather than actual value. Anything that requires upfront effort to reveal its secrets is rejected for what is fundamentally broken but covered in a veneer of user friendliness.
Python is the new Perl (Go is already gearing up to replace it where reliability is at stake) and in a couple of years the masses will look at it as a colossal and obvious failure, but they will make the same mistakes again and again when the next shiny thing comes along because the entire domain is rigged in multiple ways.
I suggest reading Richard Gabriel's essays, they are very depressing but also illuminating. In many ways he predicted what is taking place now, the rise of commodification and the obliteration of the field by companies like Google.
Before Perl, Tcl was supposed to be the Lisp replacement. As far as I know this "language X is an adequate/better replacement for Lisp" argument goes back to Ted Nelson ranting against Lisp while advocating TRAC in the 1974 book Computer Lib. So far all of the language Xs have disappeared, and the only thing to replace Lisp has been other Lisps.