Re: Google AI winner uses lisp
CL On Dec 9, 7:09 pm, Ken Wesson kwess...@gmail.com wrote: On Thu, Dec 9, 2010 at 8:53 PM, Alec Battles alec.batt...@gmail.com wrote: I'm sure a few people have read this news already. It's been up for a week, though strangely ZDnet -- which, on principle, I refuse to link to Why? -- is one of the only places to write it up. http://pr-usa.net/index.php?option=com_contenttask=viewid=560484It... Nice news to read before bedtime. Yes, but was it Clojure, or another Lisp? -- You received this message because you are subscribed to the Google Groups Clojure group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
Re: Google AI winner uses lisp
What is FFI? On Dec 9, 10:47 pm, Ken Wesson kwess...@gmail.com wrote: On Fri, Dec 10, 2010 at 12:43 AM, javajosh javaj...@gmail.com wrote: It does beg the question, though: what is a reasonable bare minimum function set that a real-life lisp would require? I think different people might give different answers to that. The academic computer scientist is likely to consider lambda, the function-call operator, nil, if, and eval to suffice. After all those suffice for anything you can express in lambda calculus. A Lisper interested in Lisp hacking as an end in itself will want to add cons, car/cdr or first/rest, =, cond, etc. The application programmer is going to additionally require FFI with, particularly, GUI libraries, not to mention various forms of disk, networking, keyboard, and mouse I/O. The systems programmer is going to additionally require being able to get at the bare metal and run tight, efficient, non-interpreted code in kernel mode. Ultimately I suppose it hinges on what someone means by a real-life Lisp. I'd say the application programmer's needs are probably th best answer there, since Lisp hacking for its own sake and academic computer science are more mathematics than real-world, and systems programming can be accomodated by a combination of C and a Lisp with some kind of FFI. (Clojure as the Lisp requires Java as well, and Clojure and C calling each other indirectly via Java and JNI as a go-between.) -- You received this message because you are subscribed to the Google Groups Clojure group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
Re: Google AI winner uses lisp
Foreign function interface. To call C/C++ libs. On Dec 10, 2010, at 6:52 PM, javajosh javaj...@gmail.com wrote: What is FFI? On Dec 9, 10:47 pm, Ken Wesson kwess...@gmail.com wrote: On Fri, Dec 10, 2010 at 12:43 AM, javajosh javaj...@gmail.com wrote: It does beg the question, though: what is a reasonable bare minimum function set that a real-life lisp would require? I think different people might give different answers to that. The academic computer scientist is likely to consider lambda, the function-call operator, nil, if, and eval to suffice. After all those suffice for anything you can express in lambda calculus. A Lisper interested in Lisp hacking as an end in itself will want to add cons, car/cdr or first/rest, =, cond, etc. The application programmer is going to additionally require FFI with, particularly, GUI libraries, not to mention various forms of disk, networking, keyboard, and mouse I/O. The systems programmer is going to additionally require being able to get at the bare metal and run tight, efficient, non-interpreted code in kernel mode. Ultimately I suppose it hinges on what someone means by a real-life Lisp. I'd say the application programmer's needs are probably th best answer there, since Lisp hacking for its own sake and academic computer science are more mathematics than real-world, and systems programming can be accomodated by a combination of C and a Lisp with some kind of FFI. (Clojure as the Lisp requires Java as well, and Clojure and C calling each other indirectly via Java and JNI as a go-between.) -- You received this message because you are subscribed to the Google Groups Clojure group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en -- You received this message because you are subscribed to the Google Groups Clojure group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
Re: Google AI winner uses lisp
On Thu, Dec 9, 2010 at 8:53 PM, Alec Battles alec.batt...@gmail.com wrote: I'm sure a few people have read this news already. It's been up for a week, though strangely ZDnet -- which, on principle, I refuse to link to Why? Because of the website's close ties to Washington. This blog post is exemplary: http://www.zdnet.com/blog/government/is-wikileaks-julian-assange-worse-than-osama-bin-laden/9578 -- You received this message because you are subscribed to the Google Groups Clojure group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
Google AI winner uses lisp
I'm sure a few people have read this news already. It's been up for a week, though strangely ZDnet -- which, on principle, I refuse to link to -- is one of the only places to write it up. http://pr-usa.net/index.php?option=com_contenttask=viewid=560484Itemid= Nice news to read before bedtime. A -- You received this message because you are subscribed to the Google Groups Clojure group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
Re: Google AI winner uses lisp
On Thu, Dec 9, 2010 at 8:53 PM, Alec Battles alec.batt...@gmail.com wrote: I'm sure a few people have read this news already. It's been up for a week, though strangely ZDnet -- which, on principle, I refuse to link to Why? -- is one of the only places to write it up. http://pr-usa.net/index.php?option=com_contenttask=viewid=560484Itemid= Nice news to read before bedtime. Yes, but was it Clojure, or another Lisp? -- You received this message because you are subscribed to the Google Groups Clojure group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
Re: Google AI winner uses lisp
Common Lisp. http://quotenil.com/ On Dec 9, 7:09 pm, Ken Wesson kwess...@gmail.com wrote: On Thu, Dec 9, 2010 at 8:53 PM, Alec Battles alec.batt...@gmail.com wrote: I'm sure a few people have read this news already. It's been up for a week, though strangely ZDnet -- which, on principle, I refuse to link to Why? -- is one of the only places to write it up. http://pr-usa.net/index.php?option=com_contenttask=viewid=560484It... Nice news to read before bedtime. Yes, but was it Clojure, or another Lisp? -- You received this message because you are subscribed to the Google Groups Clojure group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
Re: Google AI winner uses lisp
On Thu, Dec 9, 2010 at 11:55 PM, javajosh javaj...@gmail.com wrote: Common Lisp. It figures. :) -- You received this message because you are subscribed to the Google Groups Clojure group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
Re: Google AI winner uses lisp
On Dec 9, 9:07 pm, Ken Wesson kwess...@gmail.com wrote: On Thu, Dec 9, 2010 at 11:55 PM, javajosh javaj...@gmail.com wrote: Common Lisp. It figures. :) It's still a really exciting story - thanks Alec for sharing it! I was reading Gabor's post (http://quotenil.com/Planet-Wars-Post- Mortem.html) and it sounds like something I might enjoy next time. Perhaps I'll give it a shot with clojure. The exciting thing is that a) the winning program was Lisp, b) it won by a lot, and c) 99 out of the top 100 were not Lisp. Now, this could just mean that Gabor was unusually obsessed with the contest, and the language just got out of his way. Personally, I'm ok with that. -- You received this message because you are subscribed to the Google Groups Clojure group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
Re: Google AI winner uses lisp
On Fri, Dec 10, 2010 at 12:13 AM, javajosh javaj...@gmail.com wrote: On Dec 9, 9:07 pm, Ken Wesson kwess...@gmail.com wrote: On Thu, Dec 9, 2010 at 11:55 PM, javajosh javaj...@gmail.com wrote: Common Lisp. It figures. :) It's still a really exciting story - thanks Alec for sharing it! I was reading Gabor's post (http://quotenil.com/Planet-Wars-Post- Mortem.html) and it sounds like something I might enjoy next time. Perhaps I'll give it a shot with clojure. The exciting thing is that a) the winning program was Lisp, b) it won by a lot, and c) 99 out of the top 100 were not Lisp. Now, this could just mean that Gabor was unusually obsessed with the contest, and the language just got out of his way. Personally, I'm ok with that. That is indeed one of Lisp's strengths -- once you're reasonably proficient with it, the language frequently does just get out of the way, in much the way that Java doesn't. -- You received this message because you are subscribed to the Google Groups Clojure group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
Re: Google AI winner uses lisp
On Dec 9, 9:16 pm, Ken Wesson kwess...@gmail.com wrote: On Fri, Dec 10, 2010 at 12:13 AM, javajosh javaj...@gmail.com wrote: On Dec 9, 9:07 pm, Ken Wesson kwess...@gmail.com wrote: On Thu, Dec 9, 2010 at 11:55 PM, javajosh javaj...@gmail.com wrote: Common Lisp. It figures. :) It's still a really exciting story - thanks Alec for sharing it! I was reading Gabor's post (http://quotenil.com/Planet-Wars-Post- Mortem.html) and it sounds like something I might enjoy next time. Perhaps I'll give it a shot with clojure. The exciting thing is that a) the winning program was Lisp, b) it won by a lot, and c) 99 out of the top 100 were not Lisp. Now, this could just mean that Gabor was unusually obsessed with the contest, and the language just got out of his way. Personally, I'm ok with that. That is indeed one of Lisp's strengths -- once you're reasonably proficient with it, the language frequently does just get out of the way, in much the way that Java doesn't. Yeah, I just wrote in a blog comment somewhere (now lost in the aether, alas) that, basically, you can make *any* reasonable solution sketch work in a Lisp. If you want to keep thinking in Java, fine. You can do that. If you want get all functional, cool. In the end, all real programming work is just ordering the execution of imperative, side-effect generating functions. The ordering and parameterization of those calls is arbitrary (and fun!). For Clojure there's the added question of what level of side-effect functions will you want? - e.g. with javax.Swing or java.awt.Graphics? I don't know anything about Common Lisp but I'm sure they have the same decisions to make WRT C libraries. It does beg the question, though: what is a reasonable bare minimum function set that a real-life lisp would require? -- You received this message because you are subscribed to the Google Groups Clojure group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
Re: Google AI winner uses lisp
On Fri, Dec 10, 2010 at 12:43 AM, javajosh javaj...@gmail.com wrote: It does beg the question, though: what is a reasonable bare minimum function set that a real-life lisp would require? I think different people might give different answers to that. The academic computer scientist is likely to consider lambda, the function-call operator, nil, if, and eval to suffice. After all those suffice for anything you can express in lambda calculus. A Lisper interested in Lisp hacking as an end in itself will want to add cons, car/cdr or first/rest, =, cond, etc. The application programmer is going to additionally require FFI with, particularly, GUI libraries, not to mention various forms of disk, networking, keyboard, and mouse I/O. The systems programmer is going to additionally require being able to get at the bare metal and run tight, efficient, non-interpreted code in kernel mode. Ultimately I suppose it hinges on what someone means by a real-life Lisp. I'd say the application programmer's needs are probably th best answer there, since Lisp hacking for its own sake and academic computer science are more mathematics than real-world, and systems programming can be accomodated by a combination of C and a Lisp with some kind of FFI. (Clojure as the Lisp requires Java as well, and Clojure and C calling each other indirectly via Java and JNI as a go-between.) -- You received this message because you are subscribed to the Google Groups Clojure group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
