[jruby-dev] Return conversion performance revisited; passing all tests (sorta)

Attached is an updated patch for the return-value conversion/coercion optimization. It improves things significantly: ## before Measure System.currentTimeMillis, int becoming Fixnum 6.092000 0.00 6.092000 ( 6.091000) 5.929000 0.00 5.929000 ( 5.929000) 5.878000 0.00

Re: [jruby-dev] Return conversion performance revisited; passing all tests (sorta)

Charles Oliver Nutter wrote: Attached is an updated patch for the return-value conversion/coercion optimization. It improves things significantly: ## before Measure System.currentTimeMillis, int becoming Fixnum 6.092000 0.00 6.092000 ( 6.091000) 5.929000 0.00 5.929000 ( 5.

Re: [jruby-dev] Return conversion performance revisited; passing all tests (sorta)

Looks good to me. Lower was created to explicitly be able to write higher support in pure Ruby. Since the equivalent of lower has now been moved into Java, I am not sure it serves much purpose now (we do still call some methods here and there like java_object and wrap). It may be a good idea to

Re: [jruby-dev] Re: AAJUG Speakers

Nick Sieger wrote: On 8/3/07, *David Evennou* <[EMAIL PROTECTED] > wrote: Hello Nick, I am the leader of the Ann Arbor Java Users Group and am looking for presenters. Do you know anyone who can present on JRuby or any other topics at one of our m

Re: [jruby-dev] Return conversion performance revisited; passing all tests (sorta)

Charles Oliver Nutter wrote: So have a look at the contained patch. For the record, here's the failure I get running testLowerJavaSupport with this change. In my mind, the lack of consumers of "lower" support combined with the massive performance improvement make it pretty much a no-brainer to

Re: [jruby-dev] Re: AAJUG Speakers

Hi Charlie, Yes, I think that is an excellent idea! Should I coordinate with you? Thanks, David Evennou AAJUG - Original Message - From: "Charles Oliver Nutter" <[EMAIL PROTECTED]> To: Cc: "David Evennou" <[EMAIL PROTECTED]> Sent: Monday, August 06, 2007 2:39 PM Subject: Re: [jruby-d

Re: [jruby-dev] Return conversion performance revisited; passing all tests (sorta)

Charles Oliver Nutter wrote: Attached is an updated patch for the return-value conversion/coercion optimization. It improves things significantly: FYI, here's the results for a method that returns String. The improvement is about the same ratio: ## before Measure java.lang.Thread#name, Strin

Re: [jruby-dev] Bug in serialization/persistence when used with Rails

Thanks. Is the code for this Goldspike in the trunk? On Aug 2, 2007, at 4:30 PM, Robert Egglestone wrote: On every request GoldSpike selects a runtime from the pool to use for processing that request. It'll need to be updated to be compatible with your work, as it'll now need to call Ruby.se

Re: [jruby-dev] Moving to Java 5?

On 7/26/07, Charles Oliver Nutter <[EMAIL PROTECTED]> wrote: > > > It's hard for me to feel those concerns when there's multiple free Java > 5 implementations for Linux (the typical place people are running GCJ), > Sun's Java 6 version is available under very flexible distribution > licenses, and t