> Lisp was designed for single programmers managing their entire stack
Funky, McCarthy had a whole team working on and with Lisp.
Re:Why Lisp Failed (2010)
by dreamcompiler on Sunday July 07, 2019 @02:40PM
(#20376875)
So did Symbolics. And most of those programmers had full-stack mastery. That's not the case with modern software engineering projects. Nor did McCarthy or Symbolics have to deal with the manager-friendly buzzword salad that permeates most software projects today.
[Herein I mostly mean Common Lisp when I talk about Lisp. I have no experience with large projects in Clojure or Scheme.]
Most modern software -- Lisp programs included -- is written by multiple programmers in both coordinated and uncoordinated teams. Lisp can easily adapt to team programming but SWE needs to be augmented to handle programming projects that operate at a meta-level, which is natural with Lisp. If you write code at a meta level with ordinary SWE it can lead to disaster.
SWE needs tooling, procedures, and policies to deal with macros, package naming, bootstrapping, multiple-inheritance classes and first-class methods, etc. before it can really support Lisp. Lisp projects that have added such SWE features manually tend to succeed; those that don't tend to have difficulty with Lisp. My point is the team has to do the SWE work; SWE tools operating in default mode tend not to be good enough for Lisp.