Fork me on GitHub
#clojure-uk
<
2018-06-05
>
thomas06:06:16

mogge 😼

danm08:06:48

ahoy hoy

alexlynham19:06:39

still working?

korny20:06:48

on holidays - just now poking my laptop

yogidevbear20:06:48

(= (+ 👶 holiday) "Does not compute")

juxt 4
alexlynham20:06:20

@jasonbell you play the chapman stick?

alexlynham21:06:40

Quite an unusual instrument to get into

alexlynham21:06:57

Is that from an interest in prog? Jazz?

jasonbell21:06:03

@alex.lynham I went the fairly standard route of Peter Gabriel -> King Crimson -> David Sylvian 🙂

alexlynham08:06:24

ahhh, a Tony Levin fan?

jasonbell12:06:38

Yeah, but I was introduced to the instrument by a few others as well. Tony was the catalyst, in that "What the hell's that bass sound!" on I Don't Remember 🙂

alexlynham12:06:27

I think it's his birthday today, you know...

jasonbell14:06:00

aye it is, nice chap he is

jasonbell21:06:25

My friend and teacher is more the jazzer.

fmjrey22:06:36

@korny just catching up with this channel and seeing your question posted earlier this week. I'm not convinced the following would help you but just in case. Not too long ago I've worked on a command line tool to predict the impact of graphql query changes on underlying backends. The idea was to get some config data from a local file (edn and/or excel), get data from a development graphql server spitting out query results + instrumentation data such as backends being hit, query github API to retrieve queries within .graphql files, aggregate instrumentation data across these queries, all this to spit out some edn report detailing overall impact on backends and delta compared to previous run. I started with clj because of docjure (excel reporting was the initial idea). However depending on how things were going it could remain a cli only thing to be included as part of CI (in which case cljs startup time and edn are better fit), or become a constantly running service with some proper reporting UX (in which case clj and edn were fine). So I spent a lot of time converting my initial clj+excel code into cljc+edn, replacing clj only libs with cljc ones and keeping reader conditionals within lower utility namespaces. In the end I moved to a different team and stopped working on this. I did not want my cljc work to go into oblivion even for my own sake. So I extracted the more generic part into this library: https://github.com/fmjrey/cameleon I know a simple cli logic was one of your initial question. Unfortunately I did not extract the cli logic because it was part of core.cljc along with a several reader conditionals to deal with tools.cli and its cljs counterpart. But a lot of concerns regarding http io, rest io, file io, edn io, test fixtures built from http recording, http client mocking, have been extracted and could be of interest.

👍 4
korny12:06:44

I'll take a look - holidays and 👶 have restricted my fiddling time right now!

korny12:06:44
replied to a thread:@korny just catching up with this channel and seeing your question posted earlier this week. I'm not convinced the following would help you but just in case. Not too long ago I've worked on a command line tool to predict the impact of graphql query changes on underlying backends. The idea was to get some config data from a local file (edn and/or excel), get data from a development graphql server spitting out query results + instrumentation data such as backends being hit, query github API to retrieve queries within `.graphql` files, aggregate instrumentation data across these queries, all this to spit out some edn report detailing overall impact on backends and delta compared to previous run. I started with `clj` because of `docjure` (excel reporting was the initial idea). However depending on how things were going it could remain a cli only thing to be included as part of CI (in which case `cljs` startup time and edn are better fit), or become a constantly running service with some proper reporting UX (in which case `clj` and edn were fine). So I spent a lot of time converting my initial `clj`+excel code into `cljc`+edn, replacing `clj` only libs with `cljc` ones and keeping reader conditionals within lower utility namespaces. In the end I moved to a different team and stopped working on this. I did not want my `cljc` work to go into oblivion even for my own sake. So I extracted the more generic part into this library: https://github.com/fmjrey/cameleon I know a simple cli logic was one of your initial question. Unfortunately I did not extract the cli logic because it was part of `core.cljc` along with a several reader conditionals to deal with `tools.cli` and its `cljs` counterpart. But a lot of concerns regarding http io, rest io, file io, edn io, test fixtures built from http recording, http client mocking, have been extracted and could be of interest.

I'll take a look - holidays and 👶 have restricted my fiddling time right now!