[ruby-core:79845] Ruby+OMR: Status update, and future collaboration

Hi ruby-core!

I'm writing to share some updates about Ruby+OMR [1], and to start
discussing how we can collaborate a bit going forward.

Since last time I emailed, we've

* Rebased the JIT compiler to build on top of the ruby_2_4 tag [2].
* Populated the Ruby+OMR preview issue tracker, and labelled issues to
   help people navigate.
* Blogged about how the JIT compiler works [3]
* Given a talk in the Ruby DevRoom at FOSDEM 2017 [4], where we
   summarized the state of the world, and where we are going next.
* Crossed the 1000 commit threshold on the Eclipse OMR project [5], a
   nice milestone for us.

A more detailed summary of what we've been up to is included in this
blog post that I wrote [6]. I'd be honoured if you would take a couple
minutes to read that post and let me know what you think.  For those who
don't have time though, allow me to summarize briefly here:

Becoming part of the Ruby Community
===================================

The Ruby+OMR team is interested in working with the Ruby community to
make the OMR JIT compiler the default JIT compiler for CRuby. We want to
be part of making Ruby 3x3 happen. While becoming the default JIT may
not be something that will happen soon,  the most important work we have
to do is build a community interested in helping it get there. Ruby+OMR
needs community interest to succeed.

The only way we can do that, as OMR and Ruby+OMR developers, is to
provide mentorship. We are committed to helping anyone who wants to
contribute to Ruby+OMR get up and running.

We will:

* Write documentation for anything that’s unclear.
* Answer email questions.
* Be present on the Eclipse OMR channel in Slack.
* Schedule a video chat with you.
* Help you design a feature, or solve a bug.

This is the only way we can build a community, and we are dedicated to
doing it.

The Work Ahead
==============

The road forward to 3x3 performance isn't just JIT compiler work. In
order to make any JIT compiler effective, it's important to feed it a
well balanced diet of information. A good part of the work to get the
JIT compiler to help Ruby to 3x3 will be adding support for feeding the
JIT compiler information. This kind of work can be shared among any JIT
compiler that might want to become part of CRuby (which Chris Seaton
suggests may be more than one :D ), and much can be done before a JIT
compiler even gets added.

To that end, I intend to start opening some features in Redmine,
contributing some patches for features that the JIT compiler will find
useful, but that shouldn't be intrusive. I'd appreciate any and all
feedback on those patches.  I'd like to see others working on JIT
compilation technology think about sharing CRuby infrastructure as well.

If anyone is interested in helping out to make this happen, please feel
free to look at our Issue Tracker [7]  which has beginner issues and
help wanted issues tagged.

Thanks for your time,

-- Matthew Gaudet

[1]: https://github.com/rubyomr-preview/ruby
[2]: https://github.com/rubyomr-preview/ruby/commits/ruby_2_4_omr
[3]: https://developer.ibm.com/open/2016/11/18/introducing-ruby-jit/
[4]: 
https://fosdem.org/2017/schedule/event/ruby_highly_surmountable_challenges_in_ruby_omr_jit_compilation/
[5]: https://github.com/eclipse/omr
[6]: 
https://developer.ibm.com/open/2017/03/01/ruby-omr-jit-compiler-whats-next/
[7]: https://github.com/rubyomr-preview/ruby/issues


Unsubscribe: <mailto:[e-mail]?subject=unsubscribe>
<http://lists.ruby-lang.org/cgi-bin/mailman/options/ruby-core>