Fork me on GitHub

do you really need to do something like this?


while the REPL boots slowly it’s not something you’re restarting often afterwards


I generally restart it almost never


so debugging using spacemacs is a bit weird. once you hit a breakpoint you have to enter insert mode in order to enter any of the commands for the debugger (next, continue, etc)


#break and #dbg don’t seem to kick in when you load the code by evaluating the buffer


works fine by evaluating expression


@logaan: If you compile the buffer with C-c C-k, then indeed #break and #dbg are ignored. I just noticed this yesterday too, will fix it soon. For now, you canuse C-c b, which has virtually the same effect and doesn`t have that problem. (or any of the evaluation commands)


@logaan: As for spacemcas, it might be worth getting in touch with them about that. But then, if you're not in insert mode maybe the debugger keys would conflict with other global spacemacs keys. I don't know spacemacs enough to know if that's the case, but if it is that means the behavior you describe is more or less inevitable.


has anyone got figwheel working with clj-refactor? seems figwheel hard wired nrepl middlewares, and ignore nrepl-refactor middleware. wonder if there is any workaround.


clj-refactor doesn’t work with cljs anyways


all the AST-related stuff are clj dependent, although adding support for cljs is on the roadmap


@bozhidar: thanks. saved me time for digging into codes. I can forget it for now. Had thought demonstrate refactoring to my coworkers. they are kind of afraid of dynamic types, because you cannot easily debug and refactor.


do the demonstration using Clojure - by the the time they’re hooked, @benedek will add cljs support simple_smile


currently just plan to demo the reactive functional web ui with figwheel and reagent. and maybe after that, I can switch to a clojure project and demo the refactor stuff.


There's a lot of stuff that clj-refactor does entirely on the emacs side. Those parts will work with cljs.


But the heavier refactorings rely on the nrepl, so they won't


yep, that’s correct


but because @rui.yang asked about the middleware I’m guessing he’s interested in the nrepl bits