tag:blogger.com,1999:blog-5722310642266356003.post5502536440763929239..comments2024-01-07T23:21:32.676+01:00Comments on The Axis of Eval: The meaning of forms in Lisp and KernelUnknownnoreply@blogger.comBlogger4125tag:blogger.com,1999:blog-5722310642266356003.post-58756325380821301982011-08-09T03:22:35.173+02:002011-08-09T03:22:35.173+02:00Regarding fexpr use - yes, we'll have to see. ...Regarding fexpr use - yes, we'll have to see. I'm currently implementing a Kernel-like language to find out.<br /><br />Regarding compilers - fexprs and 1st-class environments definitely mess up AOT compilation completely. (You can't even do basic tricks like precompute de Bruijn indices for lexical variable lookups AOT.)<br /><br />But fexprs and 1st-class environments really are what you want from a cleanliness of design perspective. In Kernel, letrec may be faithfully defined as a macro - something that's not possible in Scheme. Furthermore, 1st-class envs do give a clean account of the top-level environment - Scheme has to use quite complicated rules to make it all work (<a href="http://www.r6rs.org/final/html/r6rs/r6rs-Z-H-13.html#node_chap_10" rel="nofollow">see here</a>).Manuel Simonihttps://www.blogger.com/profile/07840673741485280526noreply@blogger.comtag:blogger.com,1999:blog-5722310642266356003.post-59164508370421291252011-08-08T19:09:56.174+02:002011-08-08T19:09:56.174+02:00I read parts of Shutt's Kernel publications ag...I read parts of Shutt's Kernel publications again, and it seems the central issue to it all is, in a sense, hygiene -- in the sense of: how different parts of the program interact, and how that can be controlled and understood. Because that is key to what makes a language good or not: if those interrelations are harder to understand, the language will be not so good to use, and also more difficult to make good compilers/interpreters for.<br /><br />So the real question for Kernel is: is its fexpr concept good to use? We just have to see . . .Harrison Ainsworthhttp://www.hxa.name/notes/noreply@blogger.comtag:blogger.com,1999:blog-5722310642266356003.post-84339442229100129722011-08-02T17:14:00.012+02:002011-08-02T17:14:00.012+02:00Hi Gavin - same here. They're elusive.Hi Gavin - same here. They're elusive.Manuel Simonihttps://www.blogger.com/profile/07840673741485280526noreply@blogger.comtag:blogger.com,1999:blog-5722310642266356003.post-15304474386419325252011-08-02T04:45:04.470+02:002011-08-02T04:45:04.470+02:00"The Ultimate Abstraction!" This was use..."The Ultimate Abstraction!" This was useful, still trying to get a grip on what fexprs enable. I can kinda see it but then it slips away....gavinhttp://machinelake.comnoreply@blogger.com