01:50 <byteit101[m]> 2.6.10 :001 > require 'engine-hacks'... (full message at <https://libera.ems.host/_matrix/media/v3/download/libera.chat/bbad15ea49312fb5a0e8e61c2b26a80b0252ed85>)

01:51 <byteit101[m]> 2.6.10 :005 > Thread.new { p $? }

01:51 <byteit101[m]> => #<Thread:0x000055e5d32cd620@(irb):5 run>

01:51 <byteit101[m]> nil

01:51 <byteit101[m]> Yay!

17:01 <legolas_of_mirkw> Hello, I'm Martin (not Legolas 😓). I would like to ask about GSoC, specifically about the idea proposal of "JRuby Specialized String types." To get a rough initial idea, I peeked into these two classes:

17:01 <legolas_of_mirkw> https://github.com/jruby/jruby/blob/df79d1f8b874fc0710c9f21d80e23b99ca72895d/core/src/main/java/org/jruby/util/ByteList.java

17:01 <legolas_of_mirkw> https://github.com/jruby/jruby/blob/c04e08a0164885f0457197b0c4ed6acff8a3840c/core/src/main/java/org/jruby/util/RubyStringBuilder.java

17:01 <legolas_of_mirkw> However, I am still confused about what exactly the general task of "breaking our String into an abstract base and child implementations to suit different situations" means. Is there any other framework or already existing solution to this issue from which I could draw inspiration?

17:55 <byteit101[m]> You'll have to ask headius to be sure, but I'm fairly certain a perfect example of that is his exploration yesterday of >2GB strings, where large strings may want to have a different implementation than smaller strings. Unsure if rope storage would also be in there, though headius would have the plan

18:51 <byteit101[m]> Ignoring the 75% of the time race condition, I got IO.popen working in subspawn on windows!