4 ms·
Some clarification on how this differs from rbenv would be in order. Personally I'm still using rvm without problems, but I've been intrigued by the simplicity
by dasil003 15y ago
Some clarification on how this differs from rbenv would be in order. Personally I'm still using rvm without problems, but I've been intrigued by the simplicity of rbenv (especially since Bundler makes gemsets obsolete). If this is significantly simpler, how? And given that it's brand new, what kind of scenarios has it been developed and tested with?
- jayferd 15y agoGood point. A major part of the simplicity is that ry doesn't need to be loaded into your shell or shim any binaries. You simply add $PREFIX/lib/ry/current/bin to your $PATH (there's a helper to do this), and vanilla ruby/rubygems take care of the rest. Ry also punts on indexing and listing the various implemetations of Ruby out there - you give it a reference name and a source tarball url, and it runs ./configure PREFIX=...; make install. That allows it to be magic-free with respect to naming different rubies - just give it a shortcut name that you'll remember.
- callahad 15y agoIf I understand this correctly, I can only have a single Ruby active at a time, even if I have multiple terminals open, right?
- jayferd 15y agoBy default, yes. But you can also use your $PATH to select a particular ruby per session. The `fullpath` helper is for that: export PATH="$(ry fullpath mri)"