This page is not created by, affiliated with, or supported by Slack Technologies, Inc.
2019-06-26
Channels
- # announcements (6)
- # beginners (328)
- # boot (2)
- # cider (72)
- # clara (6)
- # cljdoc (4)
- # cljsrn (5)
- # clojure (78)
- # clojure-europe (3)
- # clojure-italy (22)
- # clojure-nl (4)
- # clojure-spec (3)
- # clojure-uk (114)
- # clojurescript (22)
- # clojurex (54)
- # copenhagen-clojurians (1)
- # core-async (20)
- # cursive (8)
- # data-science (1)
- # datomic (22)
- # duct (11)
- # emacs (32)
- # events (1)
- # figwheel (2)
- # fulcro (18)
- # graalvm (53)
- # graphql (39)
- # luminus (6)
- # nrepl (6)
- # off-topic (53)
- # om (1)
- # re-frame (8)
- # reagent (19)
- # reitit (3)
- # shadow-cljs (28)
- # spacemacs (10)
- # sql (37)
- # tools-deps (33)
- # vim (9)
- # xtdb (6)
Running a react-native
shadow-cljs build in Docker, and devtools on the phone is trying to connect to a ws://
URL that’s only accessible inside the Docker container (`<ws://169.254.32.96:9630/ws/...`>).
Related line in react-native.cljs
- https://github.com/thheller/shadow-cljs/blob/89acb664f800af48344288e963873bf8972ee9fb/src/main/shadow/cljs/devtools/client/react_native.cljs#L206
The ws connect fails so the app crashes with the React Native red screen every 30 seconds.
Related GH issue? https://github.com/thheller/shadow-cljs/issues/482
Are there any workarounds? Can we specify the websocket hostname in shadow-cljs.edn
?
The every-30-second crash seems to come from heartbeat!
https://github.com/thheller/shadow-cljs/blob/0e28c615cd75d4cbce27ede7a8f7f70785d42c86/src/main/shadow/cljs/devtools/client/react_native.cljs#L244
@aaron51 you can currently configure the ip/host to connect to via shadow-cljs watch app --config-merge '{:local-ip "1.2.3.4"}'
or just the :local-ip
in the build config itself
I am having some weird issue — did something change recently? ["react" :as React]
now returns an empty javascript object, and I get cascading failures about how .createContext
etc is not a function.
Trying to roll back some dependency upgrades to see if something in my package.json is playing up.
Turns out I can reproduce this after doing a release build, followed by a dev build. The dev build then breaks. Removing .shadow-cljs
fixes it. I can reproduce this 100% of the time.
if it says connected to server
on startup you did not restart the server isntance just the watch
Good point — yes, I am — I’m also nuking node_modules and .shadow_cljs
(but I’m keeping a copy around 😉 )
updated shadow-cljs version, and i had to add proxy in shadow-cljs.edn at root of project directory
sorry, false alarm, i was using the global install of shadow-cljs which was on an older version
Turns out that doing a blind npm update is dangerous. Something somewhere broke everything. Manually editing the package.json versions to the same numbers fixes the issue.
flash backs to when I was a full time JS dev
I love that whatever is in my project.clj
or deps.edn
is what I'll get. No ambiguity. The whole ^2.1.3
automatic bumping thing scared the living daylights out of me.
Not as I recall, no. It's more of a cultural problem than a software one I think? Well, I see it as a problem personally. I think lockfiles help but things can still bump mysteriously in transitive things?