I'm interested in hearing how people who use Clojure in production are debugging their code, particularly those who use Emacs.
I am having issues quickly locating problems in my Clojure code that are identified by automated integration test failures. As an example, I had a Midje test that would make an HTTP call into my application running Compojure and Liberator on Ring on Jetty and I'd get an error back in a stacktrace in my shell running the Ring process. I got to a somewhat useful error of not being able to find a terminator for a JSON object which I correctly divined was actually that the request body was empty. My usual approach to debugging these sorts of problems in C# and Visual Studio is to set the built-in debugger to break on CLR exceptions and then run the integration test with my debugger attached to the web server process and take a look at the exception and local variables starting at the bottom of the stack and going up the stack until I spot something funky. I'm completely lost in figuring out how to do something similar in Clojure and Emacs or a shell, that is to say, running something that will pause execution and let me examine the exception and in-scope variables on a stack frame by stack frame basis with source code context (bonus points if it's a REPL). One thing I've taken a look at is ritz and ritz-nrepl but I seem to get errors when I try to use them. Specifically, "Symbol's value as variable is void: nrepl-mode-map" which seems to be an already reported issue. As well, the Github repositories don't appear to have been updated in several months which makes me pessimistic that these libraries will be meaningfully maintained and as a result, I feel uncomfortable relying on them. Another thing I looked at was George Jahad's Clojure Debug Toolkit which also doesn't seem to have gotten any love in the last few years and has pretty minimal public documentation around using it.. While playing around with functions in the REPL usually does the job, sometimes the bugs and their locations are not obvious or they're at integration points where the input is a complicated object. I'd like to know what other developers who use Clojure do when they need to debug something and a REPL in a namespace with minimal execution context doesn't seem to do the trick. -- -- You received this message because you are subscribed to the Google Groups "Clojure" group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en --- You received this message because you are subscribed to the Google Groups "Clojure" group. To unsubscribe from this group and stop receiving emails from it, send an email to clojure+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/groups/opt_out.
