This page is not created by, affiliated with, or supported by Slack Technologies, Inc.
2018-07-03
Channels
- # aleph (3)
- # beginners (139)
- # boot (3)
- # cider (12)
- # cljs-dev (18)
- # clojure (100)
- # clojure-dev (21)
- # clojure-dusseldorf (5)
- # clojure-germany (1)
- # clojure-italy (35)
- # clojure-nl (26)
- # clojure-spec (4)
- # clojure-uk (60)
- # clojurescript (11)
- # clojutre (4)
- # cursive (21)
- # data-science (21)
- # datomic (47)
- # editors (3)
- # emacs (2)
- # events (4)
- # figwheel (2)
- # fulcro (28)
- # jobs (27)
- # jobs-discuss (21)
- # lein-figwheel (3)
- # midje (2)
- # off-topic (20)
- # om-next (4)
- # onyx (10)
- # overtone (1)
- # pedestal (2)
- # portkey (14)
- # re-frame (71)
- # reagent (44)
- # reitit (11)
- # remote-jobs (1)
- # ring-swagger (4)
- # shadow-cljs (64)
- # spacemacs (11)
- # testing (2)
- # tools-deps (8)
- # vim (8)
Morning.
månmån
Bore da
goooooooood morning
once I've got this serverless app I'm building done, I might try and wrap all the boilerplate (inc the repl setup etc) into a neat luminus-style template so it gives a neat bootstrapper and project structure for serverless cljs apps
the cljs-lambda stuff is cool but if I had to bring somebody else up to speed that maybe hadn't seen it before... well, it probably needs to be a bit more opinionated. Luminus level tho, not Rails-level
I've mainly used figwheel for cljs development in the browser - for cljs repls are people using that, or piggieback, or...?
norming!
@otfrom cljs for now because it's leveraging the serverless framework (http://www.serverless.com) which is node
plus no cold start penalty
@alex.lynham - Oooh, I like the look of this...
There are a bunch of use-cases in the platform I am slowly chipping away at by myself (rest assured there will be clear and active / almost boring in its repetitiveness, recruiting going on in here when I am in a position to hire people) that would really fit well into the Lambda / Serverless paradigm, and being able to implement them in cljs instead of node or Python would be dead good...
@alex.lynham - Are you using https://github.com/nervous-systems/serverless-cljs-plugin or are you doing your own thing(tm)?
@maleghast yep, I've been using that. I've pushed a couple of updates to the template. I think the maintainer wants it super-lean and unopinionated so going to work out whether the best thing is a new lein template that inherits stuff, or a fork or what
I don't want to reinvent the wheel, but I want a dev env immediately, with some sensible repl defaults and a folder structure à la luminus
it's a great plugin though
the maintainer is a legend for making it
hopefully I'll have some time this weekend to hack and think (maybe even break out the hammock if the weather holds)
@alex.lynham - Sounds great, I have to admit 🙂
what I'm trying to do is put the backend on lambda and then use re-frame for the front so it's essentially near free to host a mid-sized app with a nosql backend (dynamodb, probably)
been doing python/lambda at work and it's much more painful than my spare time cljs/serverless setup (granted, that's partly big org governance)
hmm. i wonder if i could make my cassandra client work on cljs+dynamodb... that might be interesting
> that might be interesting you can say that again
it would take some real work... the fundamental abstraction ( a cats
promise-monad ) is using manifold-deferred on clj, but works fine with bluebird promises on cljs, but there is a bunch of stuff around streaming result-sets and stream-joins which uses manifold-streams which would need porting to core.async, and i'd need to have a look at the dynamodb model to see how well the cassandra abstractions we support map on to dynamodb abstractions (e.g. LWTs for unique-key maintenance)
:wind_blowing_face: 😄 ^ that's it going right over my head. Sounds v cool though 🙂
one other point I guess about serverless is that the node vs clj (JVM) argument is a lil like python vs scala in the spark world, if you're scaling resources horizontally at the platform level, most of the work is being done at that abstraction so runtime speed doesn't matter so much maybe (although tbf node benchmarks are fast now)
another thing you might like if you are looking at serverless @alex.lynham is https://github.com/employeerepublic/deferst - it does integrant / component like stuff, but is non-blocking, i.e. supports object graphs with factory-fns delivering promises, as is common in js-land
interesting
thanks for the link!
had some time to look at the code a bit more, and actually it might be as simple as adding a profile like in luminus +piggyback
or what have you and then compiling with lumo to use the serverless-offline
node plugin
Odd question for this channel: has anyone ever owned a kitchen thermometer? We mostly cook chicken and stuff in the oven from a packet, but we had a bad experience tonight and a few other times with food that's not cooked quite right. I started looking into it, and it seems that a thermometer might be able to tell us when our chicken burger is actually cooked?
@guy am I right in saying that I'm just waiting for it to hit a particular temperature? Or am I waiting for a temperature for an amount of time?
The instant temperature that chicken is safe is around 74 C depending on the part. But you can do lower temps for longer time. 63 C sous vides is pretty awesome. See this table https://www.seriouseats.com/2015/07/the-food-lab-complete-guide-to-sous-vide-chicken-breast.html#safe
That guide is epic. I skipped this because I was cooking "Sous Vide" (which I had no idea what it was)