[jruby-dev] Java integration magic++

I just committed a patch I had lingering in JRUBY-991, plus some enhancements. We can back out the change if it causes problems or proves to be too controversial. The basic theme is auto-coercion of vanilla Ruby objects to Java interface proxies when the ruby object ends up in the same argument

[jruby-dev] [jira] Created: (JRUBY-1180) [PATCH] MacOSX: Config::CONFIG["target_os"] returns different value than MRI

[PATCH] MacOSX: Config::CONFIG["target_os"] returns different value than MRI - Key: JRUBY-1180 URL: http://jira.codehaus.org/browse/JRUBY-1180 Project: JRuby Issue

[jruby-dev] Binding scopes are specific to their containing scope

FYI, in the process of trying to eliminate bindingScope from DynamicScope, I discovered that all bindings are specific to the scope they're in. I had thought there was a single binding scope for the whole frame, but that's not the case. Witness! b = binding; eval 'x = 1', b; eval 'puts x' # =>

[jruby-dev] [jira] Created: (JRUBY-1179) Struct.new(nil, :foo) doesn't work for creating anonymous structs

Struct.new(nil, :foo) doesn't work for creating anonymous structs - Key: JRUBY-1179 URL: http://jira.codehaus.org/browse/JRUBY-1179 Project: JRuby Issue Type: Bug Comp

Re: [jruby-dev] Simplifying block yield path

Charles Oliver Nutter wrote: Attached is a patch that alters yield to always accept IRubyObject[], like a normal call, and changes all callers and implementations to handle that fact. There are ups and downs to this: How frustrating. It seems the downs are greater than I realized, and this ac

Re: [jruby-dev] Ditching SAFE and tainting

Charles Oliver Nutter-2 wrote: > > dertown wrote: >> Is there a place where we can start getting some requiremnts together for >> a >> new Ruby Security implentation. Then it would be possible to completely >> replace SAFE and Taint. >> >> Would people want to use the same api if possible?