4 ms·
From the link: The notable changes are: * VM (method cache) * RGenGC * refinements * syntax * Decimal Literal * Frozen String Literal * def's return val
by mmanfrin 13y ago
From the link:
The notable changes are:
* VM (method cache)
* RGenGC
* refinements
* syntax
* Decimal Literal
* Frozen String Literal
* def's return value
* Bignum
* 128bit
* GMP
* String#scrub
* Socket.getifaddrs
* new Rubygem
- riffraff 13y agoI think def's return value means returning symbol rather than nil? I've been waiting for it for ten years, I had almost forgotten why I wanted it in the first place :)
- _anshulk 13y agoSo, this would mean that private def foo end will now work?
- invisiblefunnel 13y agoYes: https://eval.in/49833 https://eval.in/49833
- transfire 13y agoYes. And it is somewhat surprising. Matz must have rejected this idea a dozen times. It was certainly one of the oldest feature requests around. Hard to imagine what must have changed his mind. I've rarely seem Matz change his mind.
- riffraff 13y agoI seem to recall the issue was more with returning UnboundMethod objects than with symbols, but possibly what's changed is the fact that now symbols can be GCed?
- charlysisto 13y agoPlease tell us, what's the use case ?
- bradgessler 13y agoCould somebody elaborate on "new Rubygem"? I remember reading a while back that Ruby was going to make an effort to separate stdlib components from the language, which I think is pretty exciting. The more boring side of this would be a newer version of RubyGems.
- gnufied 13y agoChangelog just means newer version of Rubygems, but with following new features: 1. Rubygems now has a dependency resolution mechanism similar to Bundler. 2. Rubygems has something called StubSpecification which means that complete gemspec of all the gems need not be loaded when using rubygems now. The idea is, we just need dependency from gemspec. Author name, URL etc are irrelevant details which used to get loaded before. So this should make it bit lightweight.
- chrislaco 13y ago"Rubygems now has a dependency resolution mechanism similar to Bundler." That's a shame, because bundler is a fail. Ok, anger aside, if I had a $ for every time json has been a dep resolution blocker on bundle install, or chef, or net-ssh, Id be rich.
- lobster_johnson 13y agoThat's not a Bundler fail, that's a gem packaging fail.
- chrislaco 13y agoYes and no. Yes, the problem is heightened by people specifying version patterns when they probably don't have to. However, when you get a bundle install conflict error about json, one gem needing a higher version, and one needing a lower version... adding json to _my_ Gemfile/gemspec to fix it when I'm not using json directly is also not the right answer. Sometimes, even that doesn't work at all.