On Sep 15, 2011, at 9:58 AM, Chad Perrin wrote:

> On Thu, Sep 15, 2011 at 10:43:23AM +0900, Aaron Jackson wrote:
>> Chris White wrote in post #1022049:
>>> 
>>> You should step outside of comfortable boundaries and experiment a
>>> little. In fact here's a 2.0 developer preview release that supports
>>> 1.9:
>>> 
>>> http://rubini.us/2011/06/07/inside-rubinius-20-preview/
>> 
>> Hmm, I like that idea too. Actually, using this method in combination
>> with the possibility to run against multiple Rubies could help track
>> down the source of problems. I do want to use 1.9 when possible.
> 
> As an addendum, I think Rubinius 1.2.4 should be "safe" to use, if you
> want to use a Rubinius version compatible with Ruby 1.8.7 -- though Chris
> White's recommendations for using 1.9.x-compatible Rubinius 2.0 make
> sense if you are willing to go that route.  I actually use RVM to test my
> code with MRI 1.8.7 and the 1.9.3 preview, as well as with Rubinius 1.2.5
> and the 2.0 preview.  My system provides an MRI Ruby 1.8.7 and an
> MRI/YARV Ruby 1.9.2, as well, but frankly I do not remember the last time
> I used either of those for anything I wrote.

FYI, the 2.0.0pre branch of Rubinius has been merged back to master. The 1.2.5 branch is available too, but everyone should be using the master branch at this point. It can be built to handle both 1.8 and 1.9 syntax (though the 1.9 syntax is incomplete).

Pop into the #rubinius channel on freenode and we'll lend you a hand getting it compiled and running.

cr