Just droppin’ this here:
;; my main work REPL:
dev=> (up-since)
#inst "2021-02-26T00:06:35.865-00:00"
;; my depstar project REPL:
dev=> (up-since)
#inst "2021-03-04T23:02:58.207-00:00"
;; my HoneySQL v2 project REPL:
user=> (dev/up-since)
#inst "2021-01-31T00:34:55.324-00:00"
(my dot-clojure
project has a :dev/repl
alias that starts a variety of REPLs and makes the up-since
function available)When I’m talking about my RDD workflow and never using refresh/reload tooling… ☝️:skin-tone-2:
I'd see it more of an evidence that startup time doesn't matter (which also is a useful point!) one can have long-running sessions with or without tools.namespace, can be considered orthogonal? my t.n sessions never become unusable (`(clear)` is a thing if things get messy)
Got addicted when I was using leiningen
💯 for the startup time issue, yeah. I don't know how long-lived REPLs are for folks that do use that refresh/reload workflow -- I don't hear them talk about that. Which is partly why I make a point of showing just how long I keep my REPLs open.
I use the refresh/reload style... my repls stay open until I need a new dependency that I cannot add with add-libs
... or if I do something stupid and lock the process up for example. I don't measure repl-uptime, and my newer (non-lein) projects start pretty quickly so this does not matter as much as it once did. I know my repls sometimes get used for many days, though.
Ah, nice to see someone else using add-libs
!
I crash things in ways that can only be fixed by restarting the repl more often than that 😅
Reminder: we have a #releases channel for releases that are just about bug fixes and/or minor enhancements, that do not belong in #announcements. Feel free to join there!