4 ms·
Reminds me of a metric the Clojure community has used in judging itself, which is "time to first XML". There's now a helpful getting started page prominently li
by Jach 7y ago
Reminds me of a metric the Clojure community has used in judging itself, which is "time to first XML". There's now a helpful getting started page prominently linked from the home page, which has no XML on it, but it used to be https://clojure.org/community/downloads https://clojure.org/community/downloads was the main page and you'll note the XML is still there. You'd see that before you even set things up, you're clearly in Java town now, buddy...
I like to use "time to first problem and the experience of its resolution" as my own ruler. Maybe it's an XML problem ("wtf is Maven?") but it's usually not until I try writing a program that I experience the first problem either at compile time or runtime. I haven't been keeping up with Clojure (whose latest version I've heard improved things on the error front) but when I first learned, it wasn't long before I made a mistake and paid for it with a giant Java stacktrace. I knew Java, though, so it wasn't a big deal to deal with it, but some instances were quite frustrating since there was a disconnect between useful messages about source locations when running code via a REPL connection versus launching the program from scratch which involves loading and running named Clojure files. It was rather subpar compared to Python, and especially to Common Lisp, which has useful compile-time warnings on top of a superior runtime error system.
Still I've always thought it'd be foolish to dive into Clojure without some Java background, or at least without the anticipation of needing to get some Java background, so it's interesting to read the contrary on this page...