hawkeyedan

I’m working on a blog post titled “which lisp” (lower case) and am soliciting responses to hopefully include in full within the post.

What do I mean by “a lisp”?

I means a lispy language.

1. S-expressions on the surface (not as a substrate or implementation detail) 2. Homoiconicity 3. Fully specified across implementations at the level of day to day use

Decision points In no particular order, here are some questions I think are relevant.

- Practicality for everyday to day generic scripting - Practicality for Web apps - Practicality for data analysis / munging tasks - Language ergonomics - Special sauce

What about Schemes?

For these purposes, each Scheme is considered a different “lisp” since in common use so many non-trivial packages/libraries/projects target a specific Scheme. Ease of learning/using other Schemes can be considered part of the special sauce, though.

What about Common Lisp?

While different CL implementations have special features, CL is fully specified and few significant packages/libraries function only on a single implementation.

What about lisp-over-another-runtime?

As long as the surface language has S-expressions and is homoiconic … it’s “a lisp” for these purposes.

Which one?

The problem domain is Lisp-y: convenient for functional programming, and made easier with macros, DSLs and/or metaprogramming.

I haven't used a proper Lisp since my programming paradigms course at Bryn Mawr many moons ago, so I don't have any strong preferences yet.

Obviously either will do the job, and Anarki is built on top of Racket.

Still, what are the pros and cons (pun intended) of each?