How Does Bundler Work, Anyway?
summary
History shows how require setup.rb and rubygems lead to bundler which does two concrete things it resolves versions ahead of time using heuristics for an NP complete dependency problem so bundle install writes a Gemfile.lock and installs an isolated app gemset and bundle exec rewrites the load path to those exact versions avoiding activation errors. Rubygems hosts ~100000 gems and ~1000000 versions. Bin stubs remove the need to type bundle exec and a 1.1 protocol change speeds installs by about 50 percent.
AI-written summary · about
chapters (15)
- 0:00Start
- 0:53@indirect
- 1:06CLOUD CITY DEVELOPMENT
- 2:25$ bundle install
- 3:11History
- 3:46it loads the codes
- 10:08ruby setup.rb setup ruby setup.rb config ruby setup.rb install
- 11:06use ruby libraries 1. Find a cool library 2. Download the library 3. Untar the library 4. Run 'ruby setup.rball
- 13:33gem install gem uninstall gem list rails
- 14:20gem install rails -v 4.1 gem install rails -v 4.2 gem install rails -v 5.0
- 14:52gem "rack", "1.0" require "rack"
- 16:32100,000 gems 1,000,000 versions
- 22:34how common can they be, really?
- 24:48the moral runtime resolution w install-time resolution
- 25:44how do we resolve at install-time?
