>the benefit came more from learning about the problem space than from switching languages. They said they had some Lisp dependencies that weren't as reliable

>the benefit came more from learning about the problem space than from switching languages. They said they had some Lisp dependencies that weren't as reliable so a hypothetical rewrite in Lisp again really doesn't change that (unless they wanted to NIH-Not-Invented-Here-rewrite the Lisp libraries themselves). Therefore, the Python ecosystem with more tested libraries was a motivation. (The Lisp dependency they don't refer to by name may have been the Lisp web server TBNL : http://edicl.github.io/hunchentoot/ http://edicl.github.io/hunchentoot/) Excerpt from: https://archive.md/Nz9Mu#selection-297.0-297.439 https://archive.md/Nz9Mu#selection-297.0-297.439 >If Lisp is so great, why did we stop using it? One of the biggest issues was the lack of widely used and tested libraries. Sure, there is a CL library for basically any task, but there is rarely more than one, and often the libraries are not widely used or well documented. Since we’re building a site largely by standing on the shoulders of others, this made things a little tougher. There just aren’t as many shoulders on which to stand.