Re: Re: JRuby on Rails

On Wednesday 22 February 2006 04:59 am, Ahmed Mohombe wrote:
> > Really it comes down to finding "small interpreter failings".  Rails
> > really stresses JRuby.  Pick a rails thing (I did *some* work toward
> > getting webrick running with an "empty app"), and run it.  When something
> > fails, figure out why by digging through the Ruby code until you can
> > narrow the failure down to a small case.  The vast majority of these turn
> > out to be easy to fix, but you'll spend hours trying to make a small
> > example of it.
>
> Hmm, isn't there a better approach (and less painful) ? E.g. using ready
> made test cases like those for Ruby itself or even RoR? - so that one
> doesn't "invent" new "small cases" but is using the existing one, thus
> being much more productive?

Running the RoR tests is not a bad idea, but the problem is (I think) that 
their "unit-tests" excercise hundreds of RoR methods (not to mention all the 
core and stdlib things).   This is probably a quick way easy way to find 
failures, but we still need to find out WHY it's failing.  In the Rails code 
this is a LOT harder than in some other code.

In a recent problem we ran across (not RoR related), there was one regex deep 
inside REXML that was causing problems.  I had a very reliable test that 
failed, but until we REALLY narrow the amount of Ruby code being 
interpretted, we can't really fix it.



-------------------------------------------------------
This SF.Net email is sponsored by xPML, a groundbreaking scripting language
that extends applications into web and mobile media. Attend the live webcast
and join the prime developer group breaking into this new coding territory!
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=110944&bid=241720&dat=121642