Can't say I'm inclined to disagree with Stallman on this one.<p>What I mean by that convuloluted expression is that while I don't understand very well the arguments for the use of cl-lib, and I'm sure it makes certain things easier, I've hated any encounter or attempt I have with using cl-lib myself. The whole thing is just weird, you're writing Emacs Lisp, but also it's not Emacs Lisp, and it behaves the same, except also different, and it litters the namespace with duplicate functionality, but it also has different naming conventions and everything behaves slightly differently.<p>Its prevalence has also led to more and more people suggesting its use (and loading) for issues that are perfectly solvable in Vanilla Elisp. Leading to more cargo culting as people struggle to build up a proper model of how the language works.<p>Again, maybe there are arguments for using it in some limited capacity at compilation time, although ideologically and aesthetically, I admit, even that seems icky to me, but I believe Stallman and others mentioned are correct in believing that something like cl-lib should not be loaded and used by something as fundamental as the inbuilt debugger if there's any way of avoiding it.