True dat..and I'd say Sun is certainly doing right by me in the performance department when it comes to 1.6:C:\jrubywork\shootout\bench\binarytrees>set JAVA_HOME=C:\j2sdk1.5.0_06C:\jrubywork\shootout\bench\binarytrees>jruby
binarytrees.ruby 11stretch tree of depth 12 check: -14096 tree
Well sun's fix for that is making low-contention locks fast and by grouping together multiple calls to the same object in one syncrhonized block (instead of multiple). i guess in mustang they will be able to remove synchronization if the compiler can prove that the Object in question is local to a
This patch is from pre-0.8.3, but was deemed too experimental to commit. I'll be committing it now (with some cleanup perhaps), but I send to the list for comments and for posterity.-- Charles Oliver Nutter @
headius.blogspot.comJRuby Developer @ jruby.sourceforge.netApplication Architect @ www.ve
It's a minor thing. I wish there was an unsynchronized Stack impl in core Java, but alas...On 4/6/06, Marc Hadley <
[EMAIL PROTECTED]> wrote:On Apr 5, 2006, at 8:14 PM, Charles O Nutter wrote:
>> - I'll be replacing Vector and Stack with ArrayList and> UnsynchronizedStack; we don't ever have more t
On 4/6/06, Marc Hadley <[EMAIL PROTECTED]> wrote:
I just did a cvs update -d but I don't see a project.xml... Netbeansdoes a decent job of integrating with an existing ant-based project,on my machine I just used the wizard and then added a new target fordebugging purposes.
I think NetBeans puts it
On Apr 5, 2006, at 8:14 PM, Charles O Nutter wrote:
- I'll be replacing Vector and Stack with ArrayList and
UnsynchronizedStack; we don't ever have more than one thread
accessing the given vars, and moving to unsynchronized collections
has given us a speed boost in many other areas.
OK,
Now you're getting deep into the bowels of my threading code. criticalizeOrDieIfKilled waits on a lock when critical= is set to true somewhere in the system. If the thread is killed, it will die once it wakes up. Both situations are handled here, and the method is called fairly frequently.
If you'r
On Apr 5, 2006, at 8:01 PM, Charles O Nutter wrote:
I saw it was a NetBeans patch, so I did what I'd been intending to
do for a while: make JRuby work as a project in NetBeans too. I
just committed project.xml, so it should be working correctly. I'm
pretty new to NetBeans, however, so if yo
Updated again:- Modified to use an inner class instead of anon, so that...- Result of executing script in-process is now saved and set/returned as appropriate for $? and result of Kernel#system callRunning Rake with this currently *does* launch scripts in process (yay!) but fails deeper in due to s
Hi.
The time has come to rewrite Zlib FAST, so it'll actually be usable. This
should not cause much of a problem and would probably speed things up a
magnitude or something like that.
I happened upon a strange error inside remote_installer.rb where RubyGems
fetches the source-index remotely:
I have too much faith in core classes being specified correctly sometimes...I'll learn someday.On 4/6/06, Nick Sieger <
[EMAIL PROTECTED]> wrote:On 4/6/06,
Charles O Nutter <[EMAIL PROTECTED]> wrote:
How's this for weird: we didn't have NameError rooted correctly.Yep, I just came to the same conc
On 4/6/06, Charles O Nutter <[EMAIL PROTECTED]> wrote:
How's this for weird: we didn't have NameError rooted correctly.Yep, I just came to the same conclusion and was about to submit a patch -- you beat me to it. Fixing the superclass for NameError fixed the example script in my first message in t
I suppose I could have just made these changes before sending the last patch...This updated version propagates the incoming streams through the interpreter, so that for MVM if you pass in in, out, and err they'll actually be used in that VM as Ruby's stdin, stdout, and stderr streams. With this add
Here's the first iteration of cleanup for Main and CommandlineParser (with Marc's Kernel changes included). I didn't make any severe structural changes, which will take more work. However, I did:- Make main() instantiate a Main instance, which now has appropriate instance vars
- Modify it to only S
Ugh. Main and CommandlineParser need some serious work for MVM to work correctly.First, they use System.exit everywhere. Error in command line params? System.exit. Unknown argument? System.exit. Error during execution (
i.e. nonzero process result)? System.exit. Double-plus ungood.Second, they use
How's this for weird: we didn't have NameError rooted correctly.In JRuby 0.8.2, 0.8.3, and HEAD we had NameError as a subclass of ScriptError, instead of StandardError. As it turns out (and I had to study a bit for this) a bare rescue automatically rescues only StandardError descendents. Since we h
16 matches
Mail list logo
