Dear Calva friends: calva https://github.com/BetterThanTomorrow/calva/releases/tag/v2.0.596
• Fix: https://github.com/BetterThanTomorrow/calva/issues/3276
• Add settings: calva.outputViews.fontSizeScale, and calva.outputViews.wordWrap
• https://github.com/BetterThanTomorrow/calva/issues/3274
Thanks @seancorfield for immediate testing the updates to the output view(s) yesterday! ❤️ 🙏 calva
Love it! Now I have wrap in editor off again and wrap in Calva output on, with the font at 0.75 of my editor. Awesome!!
So great with that immediate feedback. I could go back to my drawing board while it all was still fresh in my mind.
You know how much I like to be running the very latest versions of everything 😉
Yeah, but I also was of the mistaken idea that you were married to the terminal output. 😃 So happily surprised that I got this instant and wonderful help with the output view changes! 🙏
I like the terminal output view. Much better than the (editor) repl-output. But I always run a REPL from the command-line in a Terminal, so having to occasionally do ctrl+t ctrl+f (focus on Terminal) and then ctrl+pagedown or ctrl+pageup to switch Terminal views, just to see more details of an eval result, or driving Portal to show different views, is more work than just having it display in a sidebar.
In Cursor I can’t drag the sidebar view to below the chat, like I can with Copilot. So what I do is that I drag it to the Panel, beside the terminals. And switch order of the destinations to [sidebar, terminal]. So then I the terminal output does not pop open on connect,
Oh, that's a good trick to know... I didn't realize the order had that impact too...
I guess this is just a very flexible system now. 😃
Yeah, order determines primary destination and it is a bit special when it comes to auto-revealing, and also the reveal-output-destination command only reveals the primary one.
Awesome! Now Calva Output no longer auto-reveals, and my startup workflow is simpler! Thank you.
It’s a bit fascinating and unintuitive that so much time is spent on the output system. But then I remember that Clojurians are feedback junkies.
I think the only thing I can think of right now that would improve my workflow is for autoConnectRepl to have a variant where it only auto-connects if there is a .nrepl-port file.
Please file an issue about that. We may of course already have one.
That's because when I start VS Code, there's no REPL: I start that manually in the Terminal, and then connect. But if I reload the window, the Terminal/REPL lives on and I do want auto-reconnect then (and there's a .nrepl-port file).
Joyride can be a work-around, but it seems we could figure out a way to allow that variant too.
https://github.com/BetterThanTomorrow/calva/issues/3278
Part of the problem is that you never quite know what order the extensions are going to start up and when Calva's auto-connect will check for .nrepl-port. Occasionally, I get in fast enough—and Calva is slow enough starting up—that I get the REPL up and running before it checks and it connects seamlessly. But that is rare!
We've chatted about this before but I don't think I ever created an issue.
The #C0A8BC4TCH3 project has a fully automated connect of three different kinds of repls, including one where Joyride starts a scittle repl in a flare. https://github.com/PEZ/epupp/blob/master/.joyride/scripts/workspace_activate.cljs
My workflow across all my projects is similar, except for the exact command I use to start a REPL, which I choose depending on what I intend to do. And sometimes I do just use jack-in instead of my full dev REPL.