This page is not created by, affiliated with, or supported by Slack Technologies, Inc.
2020-04-15
Channels
- # announcements (3)
- # architecture (1)
- # babashka (52)
- # beginners (228)
- # calva (1)
- # chlorine-clover (31)
- # cider (9)
- # clj-kondo (16)
- # cljs-dev (25)
- # cljsrn (21)
- # clojure (116)
- # clojure-argentina (8)
- # clojure-europe (18)
- # clojure-france (17)
- # clojure-germany (1)
- # clojure-nl (5)
- # clojure-spec (49)
- # clojure-uk (63)
- # clojurescript (59)
- # community-development (14)
- # conjure (89)
- # core-matrix (1)
- # cursive (18)
- # data-science (1)
- # datomic (27)
- # exercism (4)
- # figwheel-main (5)
- # fulcro (38)
- # ghostwheel (8)
- # graalvm (5)
- # hoplon (2)
- # jobs-discuss (17)
- # juxt (1)
- # lambdaisland (5)
- # luminus (1)
- # lumo (9)
- # malli (7)
- # off-topic (32)
- # planck (24)
- # re-frame (14)
- # reagent (14)
- # reitit (14)
- # rum (23)
- # shadow-cljs (80)
- # spacemacs (2)
- # sql (6)
- # unrepl (1)
- # xtdb (2)
Let's add a ConjureConnect [host - optional] [port]
command since that's sorely required.
I'd like to allow people to set up mappings to connect to a staging hostname and port ideally.
Ah hah! vim.loop.getaddrinfo
! More undocumented stuff I'm discovering through example C code 😅
Ayyy
; localhost:37797 (connected)
; --------------------------------------------------------------------------------
; Assumed session: d2ec65ba-0cdc-411e-b813-2e035534de30
Now you can provide IP or host and it'll get resolved. I'll test it without an internet connection and make sure it can fallback too.
You can also change the default host from localhost just in case you're running in a docker container or something and you want your automatic .nrepl-port file connection to hook into that.
Good question! No idea yet, I'm thinking at the very least I can start the Janet process from Neovim and then read/write stdio but I'd like something a bit richer than that if I can get it.
Well yeah, maybe I'll end up with a minimal stdio based thing to begin with then upgrade to a networked (either faithful or loosely based on nREPL) implementation
Each of these will be their own challenge and tradeoffs, but should be fun since I can implement them as drop in modules that can be rewritten or replace without impacting the rest of Conjure.
how about this? https://gitlab.com/technomancy/ogion
That looks great to be honest. I wouldn't connect the Conjure Clojure nREPL client to it since it assumes Clojure (babashka is okay). So I'd extract the nREPL parts out into a reusable module and write a Racket stdio module and a Racket nREPL module I think.
As far as I can tell, if you want a good Racket eval experience you have to use DrRacket at the moment, it'll be nice to give them more options https://stackoverflow.com/questions/5352997/programming-schemeracket-with-vim-how-to-get-started
Having to edit things into the bottom of the file, save, change terminal and exec the file got boring after a while.
it's been so many years since i touched sicp, i don't remember much of what i tried with 😅
I mainly gave up because math tends to break my brain and I found that harder than the programming concepts by an order of magnitude.
I couldn't even get to the interesting "write a language" bit because I was getting stuck googling what different mathematical symbols meant (I had to draw them on a character lookup website to find out what they were called)
i have similar experiences -- but luckily these days there are more online resources and people to be easily queried
Opinions on test file running? Are these sufficient as a baseline? * run all tests * run current buffer tests * run test word under cursor? (think it'd be kind of tricky to extract the name from the root form but maybe)
And if you are in the source file, should "run this buffers tests" go and find the -test
suffix version of your namespace and test that? What if you want to test your current namespace since the tests are bundled with it.
I guess there's all, this buf and it's test counterpart (or the other way around) and root form under cursor.
:run-all-tests "ta"
:run-ns-tests "tt"
:run-current-test "tr"
Going with these as the default mappings.how about "ta" - test all, "tn" - test (in) namespace and "tc" - test (under current) cursor?
It may be technically correct, but for the user-experience, often one has to go with something that just sounds better and is easier to remember as a mnenomic.
Yes, one of the 2 classic computing problems: cache invalidation, naming and off-by-one errors.
I'm more arch linux user, but I usually check which one is active using this command:
pacmd list-sinks | grep -e 'name:' -e 'index:'
Then I can set the one I want to use with:
pacmd set-default-sink 1
where 1 is the id listed on the command above
try close the app you want to use your headphone before
Ah interesting, thanks! I'll take a look at some point. If I undock everything and just plug the jack in it's all fine
yeah, same happens here with a usb headset, so I need to do the set-default
I'm going to leave the current implementation on the legacy-jvm
branch when I'm done. Just in case people want to keep using it for a while after release.