The Little Stream Software Radar is an examination of different applied technologies and their usefulness to developers and businesses to build applications.
This serves as a mix of my preferred development stack with some foresight of how technology is changing and where new innovations might occur.
2015 Focus
Ruby and Ruby on Rails still stays a strong server-side component. I'll be focusing on the advanced architectures and quality maintenance aspects in 2015. Basically, there are a lot of good Rails applications in production with no intentions of migrating away from Ruby, while there has been an exodus to Node.js and Go.
With the rapid pace of change in client-side/frontend JavaScript, I'll be focusing on a breadth of options. The major libraries from only two years ago have fallen out of favour and it doesn't make sense to heavily invest in a single option until the community has stabilized more. Angular.js seems like a great prospect but the 1.x to 2.x path has thrown a lot of doubt into the mix for 2015.
Adopt / In production
These technologies should be adopted or are already in production. Their cost/value ratio makes them an easy benefit for any business.
- Rails 3 and 4
- Rails API for API-only applications, with an evaluation at Rails API – Tech Learning
- minitest
- Redis with an evaluation at Learning Redis
- CoffeeScript with an evaluation at Learning CoffeeScript (with a dash of Node.js)
- Knockout.js with an evaluation at Learning Knockout.js
- CSS3 animations
- QUnit with an evaluation at QUnit – Tech Learning
- Jasmine
- Puppet in a master and masterless configuration
- capistrano
Trial
These technologies are worth pursing. They will include risks but the benefits outweigh them in most cases.
- Angular.js (1.x) with an evaluation at AngularJS – Weekly Tech Learning. Slight hold due to the limited lifetime of the 1.x branch (see below).
- riot.js with an evaluation at riot.js, minimal MVP - Weekly Tech Learning
- pjax with an evaluation at Learning pjax. Turbolinks may have replaced this for new projects though.
- fog gem for EC2 and Rackspace cloud servers. An evaluation is at Fog – Automating Cloud Servers
- Padrino
- Node.js with an evaluation at Learning Node.js
- RSpec 2
- EventMachine with an evaluation at Learning EventMachine
- docker for development
Assess
These are worth exploring in order to learn and gain knowledge about where and how they can be used.
- MongoDB
- Resque
- Riak
- Cassandra
- Backbone.js with a inconclusive evaluation at Learning (and Failing at) Backbone.js
- Ember.js
- Spine.js (Build Awesome JavaScript MVC Applications)
- phantom.js (headless WebKit scriptable with a JavaScript API)
- casper.js (CasperJS is an open source navigation scripting & testing utility written in Javascript for the PhantomJS WebKit headless browser and SlimerJS (Gecko))
- Web sockets
- Socket.io
- Canvas element
- HTML5 geolocation
- VIM
- express (framework for node.js)
- tower.js (framework for node and express)
- flatiron.js (framework for node)
- derby.js (DerbyJS is a full-stack framework for writing modern web applications.)
- meteor.js
- Iridium (tutorial) (JavaScript toolchain)
- handlebars.js
- Goliath
- iOS development
- Lua
- Io
- Erlang
- spooky (Sinatra-like framework for Erlang)
- Elastic Search
- fallen (daemon library for Ruby )
- knockback
- R
- Go with a inconclusive evaluation at Go - Tech Learning
- dokku for self-hosted PaaS
- docker in a production environment
- react
- Microservices
- ECMAScript 6
- browserify
- flight.js
- lotus.rb
Hold / Sunset
Technologies on hold are ones that are too violate for evaluation, declining usability, or prior technology that is dying (sunset). Avoid relying on these on new projects and actively migrate away from on existing projects.
- Angular.js (2.x) is on hold until more information is released due to the large, backward-incompatible rewrite for 2.0.
- C will be important for closer to the metal but it's overall, practical importance for Little Stream Software is low.
Unknown
Technology that hasn't been categorized yet.
Based on and heavily influenced by ThoughtWork's Radar