Clojurians Log v2

I’ve tried clj-async-profiler, but usually I start with guesswork and sprinkling some time calls

I mostly use manual time or clj-async-profiler now for more complicated situations. I like the clarity of the report in the latter. It generally feels that CIDER features need a bit of up-front learning (keyboard shortcuts etc.) and often are so text-focused (Emacs standard). For a report I like a bit more styling and/or graphs. I do use the debugger in CIDER and sometimes wish for a slightly enhanced pretty printing experience (for a kind of watch what is going on in the app DB situation and then drill-down/explore).

I tend to use clj-async-profiler (and criterium for more fine-grained benchmarking)

I used it a few times but it has been rare. I thought it was a nice quick way to get a rough idea of functions performance

I also use clj-async-profiler and criterium instead

if cider-profile was made "pluggable" so you can delegate the sampling to a profiler of your choice (e.g. via multimethods or some other extension mechanism) it might see more use because other sampling tools are used more extensively. however, having used it a handful of times, the cider-profile-report buffer isn't much of a value-add over printing to the REPL (in one way it's worse: it gets updated in place, which can obscure the effects of changes), so it's not clear it'd be worth the effort to introduce an extension mechanism.

@afoltzm Which sampling profilers would you use if this was possible? Is it Tufte, for example?

I use Tufte and clj-async-profiler most frequently.

Evaluate the function normally.

facepalm

thanks 😄

👍 😊

All the old pages redirect so old links should still work. I don’t remember ever removing letfn from there but maybe that happened at some point