Dmitry Gutov writes: > On 10.05.2017 8:07, Gemini Lasswell wrote: > > It works well in my testing, and it avoids introducing the dynamic > variable. I'll commit it in your name (since you did most of the work > anyway) if you like it. It works in my testing too. Here's a revised version of the patch with your changes incorporated and a couple of other changes too. Since a method's argument list is supposed to be a list but not nil, I changed the test for a valid one from listp to consp. I also changed the names of method-args and edebug-match-method-args to cl-generic-method-args and edebug-match-cl-generic-method-args to better associate them with the code that uses them, and to avoid the idea that this might be new Edebug specification list functionality that should really be documented. That documentation already is complex enough that it's hard to understand and doesn't need any new complex things added to it. But if you don't like the name change feel free to change it back.