Clojurians Log v2

Seems to still work fine, but it points out that stack trace CSS might need some changes for larger displays 😄

@elken thanks for checking! And come color adjustments.

I'm overriding some of them in CSS, but yeah the stack trace parent just needs an overflow like this (maybe some padding so the scrollbar doesn't overlap)

hi @elken. Unfortunately I cannot reproduce this issue with the same clerk version. I’m causing the same exception (assoc on cons) but the error message renders as expected when the expression is evaluated. Maybe we need more context to reproduce, or the notebook source...

Sure, my repo is https://github.com/elken/aoc; I can replicate it with pretty much any notebook, I randomly picked 2023/day04 and changed one of the functions to be an invalid symbol and it triggered. Huh, interesting note that I have a block at the start of all my solution notebooks to pull the problem HTML from AOC; if I comment that out it seems to work fine...

ok, thanks!

I think we were able to reproduce the issue, thank you for reporting

Great! 😄

Not merged yet, but that’d be great if you could give "eb89e0b2e9edf8937c7c4a14c6df0ebc3217adeb" a spin and see if it still reproduces?

Sure thing!

Seems to be fixed! What was it in the end?

when catching exception we’d leave garbage in the data which we send to the client

https://github.com/nextjournal/clerk/pull/587

thanks for testing out!

Ahh, makes sense actually. Thanks for such a swift resolution 😄

@elken I think https://github.com/nextjournal/clerk/pull/588 is another take on that same problem. You’re using cider, right? Cider overrides the printing of functions to be unreadable in tools.reader based readers like edamame:

(pr-str (comp inc dec))
;; => "#function[clojure.core/comp/fn--5892]"

Not with my aoc notebooks no

ok, well anyway. Made some tweaks to the PR and merged it into main. Please let me know if it still works. 🙏

I won't be able to tonight but I will report when I can :)

no rush at all, thank you, also for the report!