Mark Jason Dominus gives a case study called “Why Lisp Will Never Win” in his “Twelve Views of Mark Jason Dominus”. He gives the following awk one-liner:

awk 'BEGIN {FS=":"}; $6=="/sbin/nologin" {print $1}' /etc/passwd

Which a member of comp.lang.lisp suggested that Lisp should implement something similar to. The response was that "with only a couple of new utility macros & functions", it could become:

(with-lines-from-file (line "/etc/passwd")
    (let ((fields (string-split line :fs #\:)))
        (when (string= (aref fields 5) "/sbin/nologin")
            (format t "~A~%" (aref fields 0))))))

Which as Dominus notes is a little over 2.5 over the length of the awk program. (but still required these macros). Naturally no one will opt to write it instead. So the problems with Common Lisp is ground-up verbosity, lack of common idioms for commonly performed tasks, and lack of motivation to use it for common, everyday (sometimes even throwaway code).

So what will it look like in Spark?

spark -inaF/:/ '(if (= (^F 5) "/sbin/login") (say (^F 0)))' /etc/passwd

(we borrowed the (array idx) notation from Arc, because an array ref object is overloaded with a method to get the index.

One can find other resources about what makes a language popular here: