[MacRuby-devel] Speed

2009-11-22 Thread Uliano Guerrini
I'm new to Ruby BECAUSE of MacRuby While learning I run this simple exercise (the sum of even fibonacci less than 4 million) on both MacRuby 0.5.2 and ruby 1.9.1. the recursive version runs about 5 times faster on MacRuby and that was expected but the iterative runs about 1000!!! times slower,

Re: [MacRuby-devel] Speed

2009-11-22 Thread Vincent Isambart
> While learning I run this simple exercise (the sum of even fibonacci less > than 4 million) on both MacRuby 0.5.2 and ruby 1.9.1. > > the recursive version runs about 5 times faster on MacRuby and that was > expected but the iterative runs about 1000!!! times slower, and that is > *frankly* u

Re: [MacRuby-devel] Speed

2009-11-22 Thread Conrad Taylor
On Sun, Nov 22, 2009 at 1:25 AM, Uliano Guerrini wrote: > I'm new to Ruby BECAUSE of MacRuby > > While learning I run this simple exercise (the sum of even fibonacci less > than 4 million) on both MacRuby 0.5.2 and ruby 1.9.1. > > the recursive version runs about 5 times faster on MacRuby and that

[MacRuby-devel] macirb fails to start on 0.5b2 (undefined method `bind' for IRB::SLex:Class)

2009-11-22 Thread Logan Bowers
Hello, I just downloaded and installed 0.5b2, however macirb fails to start; I get the following: bender:~ logan$ macirb /usr/local/bin/macirb:in `': undefined method `bind' for IRB::SLex:Class (NoMethodError) bender:~ logan$ After doing a little digging, the culprit appears to be in e2mmap

[MacRuby-devel] [MacRuby] #447: GCD Crashes

2009-11-22 Thread MacRuby
#447: GCD Crashes --+- Reporter: p...@…| Owner: lsansone...@… Type: defect| Status: new Priority: major | Milestone: MacRuby 0.5

Re: [MacRuby-devel] macirb fails to start on 0.5b2 (undefined method `bind' for IRB::SLex:Class)

2009-11-22 Thread Laurent Sansonetti
Hi Logan, Looks like a 32-bit only issue. $ arch -i386 macirb /usr/local/bin/macirb:in `': undefined method `bind' for IRB::SLex:Class (NoMethodError) Could you file a ticket on Trac about this? This way we won't forget. Thanks, Laurent On Nov 22, 2009, at 3:53 PM, Logan Bowers wrote: He

Re: [MacRuby-devel] How far away is Cucumber and RSpec support?

2009-11-22 Thread Laurent Sansonetti
Hi Giampiero, The latest 0.5 beta doesn't run rspec, AFAIK. The rspec code is using too much dark magic that MacRuby can actually handle. Last time we check, we fail to run it because our #caller returns something different than the original Ruby. We might eventually run rspec in the next

[MacRuby-devel] [MacRuby] #448: macirb fails to start on 32-bit arch (undefined method `bind' for IRB::SLex:Class)

2009-11-22 Thread MacRuby
#448: macirb fails to start on 32-bit arch (undefined method `bind' for IRB::SLex:Class) ---+ Reporter: lo...@…| Owner: lsansone...@… Type: defect | Status: new

Re: [MacRuby-devel] How far away is Cucumber and RSpec support?

2009-11-22 Thread Matt Aimonetti
If someone wants to try to run rspec 2.0 alpha, that would be cool. (code available on github) - Matt On Sun, Nov 22, 2009 at 5:15 PM, Laurent Sansonetti wrote: > Hi Giampiero, > > The latest 0.5 beta doesn't run rspec, AFAIK. The rspec code is using too > much dark magic that MacRuby can actual

Re: [MacRuby-devel] macirb fails to start on 0.5b2 (undefined method `bind' for IRB::SLex:Class)

2009-11-22 Thread Logan Bowers
Ticket #448: https://www.macruby.org/trac/ticket/448 Thanks! Logan Bowers On Nov 22, 2009, at 5:11 PM, Laurent Sansonetti wrote: > Hi Logan, > > Looks like a 32-bit only issue. > > $ arch -i386 macirb > /usr/local/bin/macirb:in `': undefined method `bind' for > IRB::SLex:Class (NoMethodErr

Re: [MacRuby-devel] Speed

2009-11-22 Thread Laurent Sansonetti
Hi Uliano, Thanks for trying MacRuby. Vincent is right. Your second test is actually running too fast (0.008167 seconds here) that most of the time is spent JIT compiling the block. Try to loop 10 times around the test and you should see that further calls are faster than Ruby 1.9.1. Or tr

Re: [MacRuby-devel] Speed

2009-11-22 Thread Matt Aimonetti
To be significant, your test should have a higher reference value. - Matt On Sun, Nov 22, 2009 at 5:30 PM, Laurent Sansonetti wrote: > Hi Uliano, > > Thanks for trying MacRuby. > > Vincent is right. Your second test is actually running too fast (0.008167 > seconds here) that most of the time is

Re: [MacRuby-devel] How far away is Cucumber and RSpec support?

2009-11-22 Thread Joshua Ballanco
I've been playing with Cucumber, but so far there are still a few issues. It doesn't seem like anything quite so hard to deal with as RSpec, but it's still early days. - Josh On Nov 22, 2009, at 8:15 PM, Laurent Sansonetti wrote: > Hi Giampiero, > > The latest 0.5 beta doesn't run rspec, AFA

Re: [MacRuby-devel] How far away is Cucumber and RSpec support?

2009-11-22 Thread Conrad Taylor
On Sun, Nov 22, 2009 at 10:08 PM, Joshua Ballanco wrote: > I've been playing with Cucumber, but so far there are still a few issues. > It doesn't seem like anything quite so hard to deal with as RSpec, but it's > still early days. > > - Josh > > Josh, do you mean you're seeing issues with Cucumbe

Re: [MacRuby-devel] Speed

2009-11-22 Thread Uliano Guerrini
Thank you all I didn't realize that JIT in macruby is compiling on the fly the code piece a piece (I thought that a source file was the compilation unit) and so JIT time to compile the block was inside my time metrics I re-run the test looping many times and the results are speaking for themsel