This page is not created by, affiliated with, or supported by Slack Technologies, Inc.
2022-06-04
Channels
- # announcements (3)
- # aws (13)
- # babashka (10)
- # beginners (30)
- # biff (1)
- # calva (59)
- # chlorine-clover (11)
- # cider (3)
- # circleci (2)
- # clojars (22)
- # clojure (21)
- # clojure-europe (2)
- # clojurescript (10)
- # core-typed (2)
- # fulcro (8)
- # girouette (12)
- # graphql (1)
- # helix (6)
- # inf-clojure (4)
- # joyride (6)
- # leiningen (7)
- # off-topic (3)
- # pathom (44)
- # polylith (13)
- # shadow-cljs (44)
- # tools-deps (1)
I've been thinking - how to keep Chlorine and Clover relevant? What is missing, what can be done for things to get better?
The thing that finally tipped me over to Calva really was inline result display, TBH.
BTW @mauricio.szabo https://marketplace.visualstudio.com/items?itemName=mauricioszabo.clover still links to the old GH repo...
Several people have asked me why I recently switched from Clover to Calva for the REPL -- Chris (Portal) asked me in a thread there about it in the context of programmatic APIs https://clojurians.slack.com/archives/C0185BFLLSE/p1651869889569069?thread_ts=1651866766.135979&cid=C0185BFLLSE
Ironically, the particular "ask" from me in that thread has been satisfied by both Clover and Calva at this point: a Joyride-enabled REPL evaluation command đŸ™‚ and it might actually be easier to work with Clover and Joyride right now because I could use Calva's nREPL connection to the Joyride server but have it send JVM evaluations via Clover's Socket REPL đŸ™‚
I think another aspect of my switch, that I haven't really articulated to anyone, is my frustration with unrepl -- I know you were talking about moving away from it but I don't know where you are on that @mauricio.szabo?
Interesting, because I'm also not satisfied with the speed that Clover is moving :(. Yeah, removing unrepl broke things. I'm at about 90% solving the issues that removing unrepl breaks, but its stikk "not there yet". Currently, I'm still assuming unrepl is on the classpath but I've removed most elisions and lazy results. The next step is to have unrepl as a possibility, not the only option.
One thing that I can do is try to hack a possibly buggy version of Clover that doesn't even have unrepl, as soon as I fix theblast issues I'm seeing, and upload the extension here. Maybe this can help identify some bugs...
I can branch my VS Code config to have a Clover-based branch again for testing...
@U04V70XH6 just to not let this hanging, I'll focus my efforts on other things right now that Microsoft announced the official end of Atom, so these features (avoid UNREPL, inline results in Clover, etc) will not be on my recent roadmap either... so don't branch anything, keep yourself in Calva - I'll work together with then if they decide to have a pure socket-repl evaluation on Calva
That's fair enough. I thought about you and Clover when I saw the EOL announcement... It's not exactly a surprise.