Comments on: How Rubinius SendSites Work – Part 1 http://betterruby.wordpress.com/2008/03/19/how-rubinius-sendsites-work-part-1/ Covering the developmnet of Rubinius, a new Ruby implementation and virtual machine Sun, 29 Mar 2009 17:59:08 +0000 http://wordpress.com/ hourly 1 By: -= Linkage 2008.03.24 =- http://betterruby.wordpress.com/2008/03/19/how-rubinius-sendsites-work-part-1/#comment-68 -= Linkage 2008.03.24 =- Mon, 26 Jan 2009 15:37:02 +0000 http://betterruby.wordpress.com/?p=3#comment-68 [...] How Rubinius SendSites Work - Part 1 [...] [...] How Rubinius SendSites Work – Part 1 [...]

]]> By: Nome do Jogo » Blog Archive » Rails Podcast Brasil - Episódio 10 http://betterruby.wordpress.com/2008/03/19/how-rubinius-sendsites-work-part-1/#comment-66 Nome do Jogo » Blog Archive » Rails Podcast Brasil - Episódio 10 Mon, 25 Aug 2008 13:18:34 +0000 http://betterruby.wordpress.com/?p=3#comment-66 [...] How Rubinius SendSites Work – Part 1 [...] [...] How Rubinius SendSites Work – Part 1 [...]

]]> By: agardiner http://betterruby.wordpress.com/2008/03/19/how-rubinius-sendsites-work-part-1/#comment-27 agardiner Tue, 01 Apr 2008 04:35:52 +0000 http://betterruby.wordpress.com/?p=3#comment-27 Ola: You're right, and CallSite was certainly considered as the class name. However, I believe one of the reasons Evan settled on the name SendSite is that it is used as an argument to opcodes that all start with send_. Ola: You’re right, and CallSite was certainly considered as the class name. However, I believe one of the reasons Evan settled on the name SendSite is that it is used as an argument to opcodes that all start with send_.

]]> By: How Rubinius SendSites Work - Part 2 « Building a Better Ruby http://betterruby.wordpress.com/2008/03/19/how-rubinius-sendsites-work-part-1/#comment-25 How Rubinius SendSites Work - Part 2 « Building a Better Ruby Tue, 01 Apr 2008 04:28:12 +0000 http://betterruby.wordpress.com/?p=3#comment-25 [...] Building a Better Ruby Covering the developmnet of Rubinius, a new Ruby implementation and virtual machine « How Rubinius SendSites Work - Part 1 [...] [...] Building a Better Ruby Covering the developmnet of Rubinius, a new Ruby implementation and virtual machine « How Rubinius SendSites Work – Part 1 [...]

]]> By: Sp3w » Blog Archive » Linkage 2008.03.24 http://betterruby.wordpress.com/2008/03/19/how-rubinius-sendsites-work-part-1/#comment-22 Sp3w » Blog Archive » Linkage 2008.03.24 Mon, 24 Mar 2008 18:35:48 +0000 http://betterruby.wordpress.com/?p=3#comment-22 [...] How Rubinius SendSites Work - Part 1 [...] [...] How Rubinius SendSites Work – Part 1 [...]

]]> By: Ola Bini http://betterruby.wordpress.com/2008/03/19/how-rubinius-sendsites-work-part-1/#comment-15 Ola Bini Thu, 20 Mar 2008 12:43:12 +0000 http://betterruby.wordpress.com/?p=3#comment-15 Not sure why you're using the word SendSite for something that is in all the literature called a CallSite. I realize that you want to emphasize the message passing idiom, but even most SmallTalk literature call these structures CallSites. Josh: CallSite's allow inline caches, but are not the same thing. Also note that this CallSite implementation is more like a monomorphic inline cache at the moment, rather than a polymorphic one. Not sure why you’re using the word SendSite for something that is in all the literature called a CallSite. I realize that you want to emphasize the message passing idiom, but even most SmallTalk literature call these structures CallSites.

Josh: CallSite’s allow inline caches, but are not the same thing. Also note that this CallSite implementation is more like a monomorphic inline cache at the moment, rather than a polymorphic one.

]]> By: agardiner http://betterruby.wordpress.com/2008/03/19/how-rubinius-sendsites-work-part-1/#comment-5 agardiner Wed, 19 Mar 2008 22:24:49 +0000 http://betterruby.wordpress.com/?p=3#comment-5 crayz: Links to the source is a great idea, I'll do that from now on. josh: A sendsite lookup function that uses polymorphic inline caching is planned; that's the great thing about send sites - the flexibility they provide allows different approaches to be taken within the same framework; I plan to cover this in part 2... Cheers, Adam crayz: Links to the source is a great idea, I’ll do that from now on.

josh: A sendsite lookup function that uses polymorphic inline caching is planned; that’s the great thing about send sites – the flexibility they provide allows different approaches to be taken within the same framework; I plan to cover this in part 2…

Cheers,

Adam

]]> By: josh http://betterruby.wordpress.com/2008/03/19/how-rubinius-sendsites-work-part-1/#comment-4 josh Wed, 19 Mar 2008 15:37:11 +0000 http://betterruby.wordpress.com/?p=3#comment-4 This is cool stuff. Thanks for the nice writeup. SendSites look like an interesting variation on polymorphic inline method caches. The Selector class in particular seems useful. Runtime performance and metrics aside, that sort of information can be very useful for building an IDE with refactoring capabilities. Glad to see that coming to Ruby. This is cool stuff. Thanks for the nice writeup.

SendSites look like an interesting variation on polymorphic inline method caches. The Selector class in particular seems useful. Runtime performance and metrics aside, that sort of information can be very useful for building an IDE with refactoring capabilities. Glad to see that coming to Ruby.

]]> By: crayz http://betterruby.wordpress.com/2008/03/19/how-rubinius-sendsites-work-part-1/#comment-3 crayz Wed, 19 Mar 2008 14:06:45 +0000 http://betterruby.wordpress.com/?p=3#comment-3 Thanks, it's really nice to see you guys blog about this stuff. One comment - it would be nice to provide links to gitweb for the source of the classes you mention Thanks, it’s really nice to see you guys blog about this stuff. One comment – it would be nice to provide links to gitweb for the source of the classes you mention

]]>