Console is a web app that allows you to query Crux and see data from different perpectives.

It can be used as a standalone analytics tool or as an educational tool.

Console comes with a set of example queries but you can make it load your own set by providing a url param examples-gist=gist-raw-url, like here This way Console will import this file. Important note : link must point to raw gist content. Preset queries example:

[{:title "Space Stocks"
  :query {:find [e]
          :where [[e :crux.db/id _]]}}
 {:title "Add space stocks"
  :query [[:crux.tx/put {:crux.db/id 33 :title "Company 3"}]]}]

Can be used as an analytics tool For nginx – the following redirects may be used

# in a server block
location /console {                                                                                                                                            
        # don't forget about at least basic auth
        proxy_pass http://localhost:5000;                                                                                                                      
}                                                                                                                                                              
location /crux {                                                                                                                                               
        rewrite ^(/crux)(/.*)$ $2 break;                                                                                                                       
        proxy_pass http://localhost:4889/;
}

--frontend-port  5000
--embed-crux     false
--conf-file      crux-console-conf.edn
--crux-http-port 8080

All builds require node and yarn to be installed.

yarn is, at the moment, a much better alternative to npm. Yarn ROI is that in 2-3 uses it will save more time than you spent installing and learning it, compared to 2-3 runs of npm install.

1. lein build will produce you production ready assets
2. lein build-ebs does the build above and packs it into Elastic Beanstalk package which we upload to AWS to see console.crux.cloud

To launch development REPL

lein cljs-dev  # will install all the node modules, launch cljs compiling guard with code hotswapping
lein repl      # will launch edge rebel repl

# once
yarn install

dev/shadow-dev &
# will launch shadow-cljs watch build
# it runs a local version

lein repl 
# will launch the server

Note that Plotly has its own packaging model (due to heavy weight beneath), so search for different packages on npm if you need more fine-grained control. e.g.

• All-in-one https://www.npmjs.com/package/plotly.js-dist

There are a few things we can look at

• Pikaday

  • looking nice and simple
  • no deps
  • url https://github.com/timgilbert/cljs-pikaday
  • no timepicker, but that's an easier fix

• React datetime picker

  • demo url http://projects.wojtekmaj.pl/react-datetime-picker/
  • gh url: https://github.com/wojtekmaj/react-datetime-picker
  • good UX, has timepicker
  • claimed test cov 92%
  • seemingly lots of deps :(

• Hacker one datepicker

  • demo link: https://reactdatepicker.com
  • gh link: https://github.com/Hacker0x01/react-datepicker/
  • Lots of options, arguable best datepicker-timepicker combination
  • quite heavy :(

Things to be aware of

• YouCanBookMe/react-datetime
  • drags moment.js as a dep
  • UX is questionable

Can be improved with skipping with uberjar build and just updating the jar with

jar uf jar-file input-file(s)