tag:blogger.com,1999:blog-2447174102813539049.post6078982268712772069..comments2024-03-29T03:43:02.688-07:00Comments on Room 101: Patterns as Objects in NewspeakGilad Brachahttp://www.blogger.com/profile/17934280339206214042noreply@blogger.comBlogger7125tag:blogger.com,1999:blog-2447174102813539049.post-73259466716681249122010-07-19T18:25:56.444-07:002010-07-19T18:25:56.444-07:00minobuzo: I assume you are asking whether I can ov...minobuzo: I assume you are asking whether I can override the classes used to create the pattern literals?<br />Pattern literals are translated into calls of the form Pattern literal:, Pattern wildcard etc. Since Pattern si just a message send, you can override it to return a different class. Down th eline, the goal is to compile all literals this way.Gilad Brachahttps://www.blogger.com/profile/17934280339206214042noreply@blogger.comtag:blogger.com,1999:blog-2447174102813539049.post-21797451884583035752010-07-19T18:13:06.075-07:002010-07-19T18:13:06.075-07:00This looks really interesting! I'm a follower ...This looks really interesting! I'm a follower of your blog and my question is regarding literals (in this case the pattern objects) and your modular idea? How can you specified the classes involved in the compilation?Anonymoushttps://www.blogger.com/profile/06971928468365151728noreply@blogger.comtag:blogger.com,1999:blog-2447174102813539049.post-28411252009222427492010-06-28T05:51:18.488-07:002010-06-28T05:51:18.488-07:00Mark,
Thanks. One could probably live for a long ...Mark,<br /><br />Thanks. One could probably live for a long time without this extension, but people I respect are quite adamant about the advantages. For now, it is just an experiment. The details will be adjusted and we will decide over time.Gilad Brachahttps://www.blogger.com/profile/17934280339206214042noreply@blogger.comtag:blogger.com,1999:blog-2447174102813539049.post-54103996349275316022010-06-27T19:35:21.556-07:002010-06-27T19:35:21.556-07:00Raffello:
Your point is well taken, and I am keen...Raffello:<br /><br />Your point is well taken, and I am keenly aware of it. The extension is experimental and the rules may change. They certainly need to be pinned down.<br /><br />Clearly, lexically visible names must have priority (this isn't really the case in the current implementation). One might even argue against playing any dynamic scope games, and requiring the closure passed to <= to take an explicit argument if access to the binding is needed.<br /><br />In short, more experience is needed.Gilad Brachahttps://www.blogger.com/profile/17934280339206214042noreply@blogger.comtag:blogger.com,1999:blog-2447174102813539049.post-75801362883690086452010-06-27T03:37:46.016-07:002010-06-27T03:37:46.016-07:00Cute Idea :), although I wonder if the sugar is re...Cute Idea :), although I wonder if the sugar is really needed. Please pass my congradulations onto Felix if/when you see him Gilad. As always, I look forward to playing with the next version of Newspeak.Mark Lee Smithhttps://www.blogger.com/profile/12819093371910833234noreply@blogger.comtag:blogger.com,1999:blog-2447174102813539049.post-64442233517174598382010-06-19T10:49:03.533-07:002010-06-19T10:49:03.533-07:00Thanks. Fixed now.Thanks. Fixed now.Gilad Brachahttps://www.blogger.com/profile/17934280339206214042noreply@blogger.comtag:blogger.com,1999:blog-2447174102813539049.post-29486490166397853762010-06-19T07:17:02.722-07:002010-06-19T07:17:02.722-07:00The <...> tags in the post seem broken. It k...The <...> tags in the post seem broken. It keeps saying things like, "Instances of Product will match patterns of the form <> for some x and y.". Surely there should be something between the <...>.Daniel Armakhttps://www.blogger.com/profile/02908694307839822186noreply@blogger.com