Re: [fonc] yet another meta compiler compiler
another meta compiler compiler While we're all waiting for the next STEP report I thought I'd share something I've been working on, inspired by O'Meta and by the Meta-II paper and by the discussions on this list from November. I've written up the construction of a parser generator and compiler compiler here: https://github.com/charlesap/ibnf/blob/master/SyntaxChapter.pdf?raw=true The source code can be had here: https://github.com/charlesap/ibnf Don't be fooled by the footnotes and references, this is a piece of outsider literature. I am a barbarian come to pillage the ivory tower. Yarr. Chuck ___ fonc mailing list fonc@vpri.org http://vpri.org/mailman/listinfo/fonc ___ fonc mailing list fonc@vpri.org http://vpri.org/mailman/listinfo/fonc ___ fonc mailing list fonc@vpri.org http://vpri.org/mailman/listinfo/fonc ___ fonc mailing list fonc@vpri.org http://vpri.org/mailman/listinfo/fonc ___ fonc mailing list fonc@vpri.org http://vpri.org/mailman/listinfo/fonc ___ fonc mailing list fonc@vpri.org http://vpri.org/mailman/listinfo/fonc
Re: [fonc] yet another meta compiler compiler
What will this book be called? I would like to get one when it comes out. Why do you always start with C? Why not start with a higher language like Smalltalk? You could compile ByteArrays that could then run fast. Wouldn't it be more self documenting? You would have access to all the Smalltalk resources. Because C is the portable assembler. Because C has better access to all of the OS and all other languages. By starting with C you leverage the C compiler. You would have to write a C compiler in Smalltalk to get the same thing. I am writing BorgLisp in Smalltalk which is supposed to asimilate all other computer languages into itself as dialects of lisp. Like Clojure asimilates Java into itself. Like Cola was a combination of Lisp and Smalltalk and C. I have one dialect of Lisp in BorgLisp so far and a source level stepping debugger that can handle macros( it should ). It's written in Dolphin Smalltalk but I would like to port it over to Pharo or VisualWorks. If I tried to do the same thing in C I don't think I could. Please explain why starting with C is better than this. Is there a debugger for C that is as good as the one in Smalltalk? Do you use Test Driven Development as a way to get around not having a debugger? What programming environment do you use? I look at Maru and there are absolutely no comments in there. A litterate version of Maru would be way too cool to ever actually happen in this cursed universe we live in. I hope Maru is what I'm talking about but I can't remember if it is or not. I hope you write this book with the literate meta compiler compiler in it. I hope I get my hands on it. I wish Maru was literate. Now what Favorite shall I put this link under so that I might have some chance of ever seeing it again. I suppose my Git account would be a good place to do that. But I don't Git over there that much. If I seem scattered maybe it's because I am a high functioning autistic. He said. On Fri, Feb 8, 2013 at 8:08 PM, Charles Perkins ch...@kuracali.com wrote: Thank you, Alan! That means quite a bit to me. I was so looking forward to a good pillage. Ah well… to follow your metaphor (a kinder one and more inclusive to be sure) I am most appreciative of the generous sharing of seeds going on around here. Chuck On Feb 8, 2013, at 5:42 PM, Alan Kay alan.n...@yahoo.com wrote: Looks nice to me! But no ivory towers around to pillage. (However planting a few seeds is almost always a good idea) Cheers, Alan From: Charles Perkins ch...@kuracali.com To: Fundamentals of New Computing fonc@vpri.org Sent: Friday, February 8, 2013 3:52 PM Subject: [fonc] yet another meta compiler compiler While we're all waiting for the next STEP report I thought I'd share something I've been working on, inspired by O'Meta and by the Meta-II paper and by the discussions on this list from November. I've written up the construction of a parser generator and compiler compiler here: https://github.com/charlesap/ibnf/blob/master/SyntaxChapter.pdf?raw=true The source code can be had here: https://github.com/charlesap/ibnf Don't be fooled by the footnotes and references, this is a piece of outsider literature. I am a barbarian come to pillage the ivory tower. Yarr. Chuck ___ fonc mailing list fonc@vpri.org http://vpri.org/mailman/listinfo/fonc ___ fonc mailing list fonc@vpri.org http://vpri.org/mailman/listinfo/fonc ___ fonc mailing list fonc@vpri.org http://vpri.org/mailman/listinfo/fonc ___ fonc mailing list fonc@vpri.org http://vpri.org/mailman/listinfo/fonc
Re: [fonc] yet another meta compiler compiler
C? Where? I do admit to using Python for semantic productions. Python is terse and it was helpful to be able to pass tuples around. That said, I do not think that C is bad. One nice thing about C is its very lack of a helpful runtime. its up to you to supply that however you want. That is of course also its drawback. I think Maru is awesome, I'm still wrapping my head around COLAs. Every language is literate when embedded in a versatile markup… Thank you for taking a look and sharing your observations. Perhaps one day I will finish this book and give it a title. That is my intention, but you know about intentions... Meanwhile, Best wishes, Chuck On Feb 9, 2013, at 11:27 AM, Kjell Godo squeakl...@gmail.com wrote: What will this book be called? I would like to get one when it comes out. Why do you always start with C? Why not start with a higher language like Smalltalk? You could compile ByteArrays that could then run fast. Wouldn't it be more self documenting? You would have access to all the Smalltalk resources. Because C is the portable assembler. Because C has better access to all of the OS and all other languages. By starting with C you leverage the C compiler. You would have to write a C compiler in Smalltalk to get the same thing. I am writing BorgLisp in Smalltalk which is supposed to asimilate all other computer languages into itself as dialects of lisp. Like Clojure asimilates Java into itself. Like Cola was a combination of Lisp and Smalltalk and C. I have one dialect of Lisp in BorgLisp so far and a source level stepping debugger that can handle macros( it should ). It's written in Dolphin Smalltalk but I would like to port it over to Pharo or VisualWorks. If I tried to do the same thing in C I don't think I could. Please explain why starting with C is better than this. Is there a debugger for C that is as good as the one in Smalltalk? Do you use Test Driven Development as a way to get around not having a debugger? What programming environment do you use? I look at Maru and there are absolutely no comments in there. A litterate version of Maru would be way too cool to ever actually happen in this cursed universe we live in. I hope Maru is what I'm talking about but I can't remember if it is or not. I hope you write this book with the literate meta compiler compiler in it. I hope I get my hands on it. I wish Maru was literate. Now what Favorite shall I put this link under so that I might have some chance of ever seeing it again. I suppose my Git account would be a good place to do that. But I don't Git over there that much. If I seem scattered maybe it's because I am a high functioning autistic. He said. On Fri, Feb 8, 2013 at 8:08 PM, Charles Perkins ch...@kuracali.com wrote: Thank you, Alan! That means quite a bit to me. I was so looking forward to a good pillage. Ah well… to follow your metaphor (a kinder one and more inclusive to be sure) I am most appreciative of the generous sharing of seeds going on around here. Chuck On Feb 8, 2013, at 5:42 PM, Alan Kay alan.n...@yahoo.com wrote: Looks nice to me! But no ivory towers around to pillage. (However planting a few seeds is almost always a good idea) Cheers, Alan From: Charles Perkins ch...@kuracali.com To: Fundamentals of New Computing fonc@vpri.org Sent: Friday, February 8, 2013 3:52 PM Subject: [fonc] yet another meta compiler compiler While we're all waiting for the next STEP report I thought I'd share something I've been working on, inspired by O'Meta and by the Meta-II paper and by the discussions on this list from November. I've written up the construction of a parser generator and compiler compiler here: https://github.com/charlesap/ibnf/blob/master/SyntaxChapter.pdf?raw=true The source code can be had here: https://github.com/charlesap/ibnf Don't be fooled by the footnotes and references, this is a piece of outsider literature. I am a barbarian come to pillage the ivory tower. Yarr. Chuck ___ fonc mailing list fonc@vpri.org http://vpri.org/mailman/listinfo/fonc ___ fonc mailing list fonc@vpri.org http://vpri.org/mailman/listinfo/fonc ___ fonc mailing list fonc@vpri.org http://vpri.org/mailman/listinfo/fonc ___ fonc mailing list fonc@vpri.org http://vpri.org/mailman/listinfo/fonc ___ fonc mailing list fonc@vpri.org http://vpri.org/mailman/listinfo/fonc
Re: [fonc] yet another meta compiler compiler
Borglisp, hey? Sounds like Ometa to me. Julian On 10/02/2013, at 6:27 AM, Kjell Godo squeakl...@gmail.com wrote: What will this book be called? I would like to get one when it comes out. Why do you always start with C? Why not start with a higher language like Smalltalk? You could compile ByteArrays that could then run fast. Wouldn't it be more self documenting? You would have access to all the Smalltalk resources. Because C is the portable assembler. Because C has better access to all of the OS and all other languages. By starting with C you leverage the C compiler. You would have to write a C compiler in Smalltalk to get the same thing. I am writing BorgLisp in Smalltalk which is supposed to asimilate all other computer languages into itself as dialects of lisp. Like Clojure asimilates Java into itself. Like Cola was a combination of Lisp and Smalltalk and C. I have one dialect of Lisp in BorgLisp so far and a source level stepping debugger that can handle macros( it should ). It's written in Dolphin Smalltalk but I would like to port it over to Pharo or VisualWorks. If I tried to do the same thing in C I don't think I could. Please explain why starting with C is better than this. Is there a debugger for C that is as good as the one in Smalltalk? Do you use Test Driven Development as a way to get around not having a debugger? What programming environment do you use? I look at Maru and there are absolutely no comments in there. A litterate version of Maru would be way too cool to ever actually happen in this cursed universe we live in. I hope Maru is what I'm talking about but I can't remember if it is or not. I hope you write this book with the literate meta compiler compiler in it. I hope I get my hands on it. I wish Maru was literate. Now what Favorite shall I put this link under so that I might have some chance of ever seeing it again. I suppose my Git account would be a good place to do that. But I don't Git over there that much. If I seem scattered maybe it's because I am a high functioning autistic. He said. On Fri, Feb 8, 2013 at 8:08 PM, Charles Perkins ch...@kuracali.com wrote: Thank you, Alan! That means quite a bit to me. I was so looking forward to a good pillage. Ah well… to follow your metaphor (a kinder one and more inclusive to be sure) I am most appreciative of the generous sharing of seeds going on around here. Chuck On Feb 8, 2013, at 5:42 PM, Alan Kay alan.n...@yahoo.com wrote: Looks nice to me! But no ivory towers around to pillage. (However planting a few seeds is almost always a good idea) Cheers, Alan From: Charles Perkins ch...@kuracali.com To: Fundamentals of New Computing fonc@vpri.org Sent: Friday, February 8, 2013 3:52 PM Subject: [fonc] yet another meta compiler compiler While we're all waiting for the next STEP report I thought I'd share something I've been working on, inspired by O'Meta and by the Meta-II paper and by the discussions on this list from November. I've written up the construction of a parser generator and compiler compiler here: https://github.com/charlesap/ibnf/blob/master/SyntaxChapter.pdf?raw=true The source code can be had here: https://github.com/charlesap/ibnf Don't be fooled by the footnotes and references, this is a piece of outsider literature. I am a barbarian come to pillage the ivory tower. Yarr. Chuck ___ fonc mailing list fonc@vpri.org http://vpri.org/mailman/listinfo/fonc ___ fonc mailing list fonc@vpri.org http://vpri.org/mailman/listinfo/fonc ___ fonc mailing list fonc@vpri.org http://vpri.org/mailman/listinfo/fonc ___ fonc mailing list fonc@vpri.org http://vpri.org/mailman/listinfo/fonc ___ fonc mailing list fonc@vpri.org http://vpri.org/mailman/listinfo/fonc
Re: [fonc] yet another meta compiler compiler
Where are the other chapters and what is the book about? On Fri, Feb 8, 2013 at 6:52 PM, Charles Perkins ch...@kuracali.com wrote: While we're all waiting for the next STEP report I thought I'd share something I've been working on, inspired by O'Meta and by the Meta-II paper and by the discussions on this list from November. I've written up the construction of a parser generator and compiler compiler here: https://github.com/charlesap/ibnf/blob/master/SyntaxChapter.pdf?raw=true The source code can be had here: https://github.com/charlesap/ibnf Don't be fooled by the footnotes and references, this is a piece of outsider literature. I am a barbarian come to pillage the ivory tower. Yarr. Chuck ___ fonc mailing list fonc@vpri.org http://vpri.org/mailman/listinfo/fonc -- (display hello world) ; Scheme console.log(hello world); // JavaScript ___ fonc mailing list fonc@vpri.org http://vpri.org/mailman/listinfo/fonc
Re: [fonc] yet another meta compiler compiler
Looks nice to me! But no ivory towers around to pillage. (However planting a few seeds is almost always a good idea) Cheers, Alan From: Charles Perkins ch...@kuracali.com To: Fundamentals of New Computing fonc@vpri.org Sent: Friday, February 8, 2013 3:52 PM Subject: [fonc] yet another meta compiler compiler While we're all waiting for the next STEP report I thought I'd share something I've been working on, inspired by O'Meta and by the Meta-II paper and by the discussions on this list from November. I've written up the construction of a parser generator and compiler compiler here: https://github.com/charlesap/ibnf/blob/master/SyntaxChapter.pdf?raw=true The source code can be had here: https://github.com/charlesap/ibnf Don't be fooled by the footnotes and references, this is a piece of outsider literature. I am a barbarian come to pillage the ivory tower. Yarr. Chuck ___ fonc mailing list fonc@vpri.org http://vpri.org/mailman/listinfo/fonc ___ fonc mailing list fonc@vpri.org http://vpri.org/mailman/listinfo/fonc
Re: [fonc] yet another meta compiler compiler
Er… Hmm… About that… In the (missing) Chapter 1 I attempt to illustrate building a computation model using a declarative and incremental programming language. Towards the end of that effort I realized I really needed an implementation of the language. I really like Alessandro's work but when I saw the Meta II paper I saw the way I wanted to do it in at least this first implementation. A proper conclusion to Chapter 2 would come in three parts: implement in the metacompiler (e.g. ibnf/six) a compiler for a minimal but general purpose language, then b) reimplement the semantics for the language in its own self, then c) reimplement the semantics for the metacompiler in that language. At that point a self-sustaining development environment would be established. If each of a, b and c can be done in 10 pages or so of literate programming text, I think that would be pretty cool. It would be especially cool if at that point it was easy to extend the system to integrate the source code and programming styles of other systems. Where there are two chapters there are sure to be more. I would like to continue with an operating system kernel or two, networking, graphics, text processing and the manipulation and distributed storage of relational data. Wirth and Gutknect's Project Oberon is my inspiration for the project and STEPs shows another way to build such a system. My own way would of course be a third path to the peak of the same mountain of a self-sustaining computing system, succinctly and fully described, useful for every-day tasks. Its a Quixotic quest but I think the best quests are. Thanks for taking a look! Chuck On Feb 8, 2013, at 4:48 PM, Matthew Retchin mretc...@gmail.com wrote: Where are the other chapters and what is the book about? On Fri, Feb 8, 2013 at 6:52 PM, Charles Perkins ch...@kuracali.com wrote: While we're all waiting for the next STEP report I thought I'd share something I've been working on, inspired by O'Meta and by the Meta-II paper and by the discussions on this list from November. I've written up the construction of a parser generator and compiler compiler here: https://github.com/charlesap/ibnf/blob/master/SyntaxChapter.pdf?raw=true The source code can be had here: https://github.com/charlesap/ibnf Don't be fooled by the footnotes and references, this is a piece of outsider literature. I am a barbarian come to pillage the ivory tower. Yarr. Chuck ___ fonc mailing list fonc@vpri.org http://vpri.org/mailman/listinfo/fonc -- (display hello world) ; Scheme console.log(hello world); // JavaScript ___ fonc mailing list fonc@vpri.org http://vpri.org/mailman/listinfo/fonc ___ fonc mailing list fonc@vpri.org http://vpri.org/mailman/listinfo/fonc
Re: [fonc] yet another meta compiler compiler
Thank you, Alan! That means quite a bit to me. I was so looking forward to a good pillage. Ah well… to follow your metaphor (a kinder one and more inclusive to be sure) I am most appreciative of the generous sharing of seeds going on around here. Chuck On Feb 8, 2013, at 5:42 PM, Alan Kay alan.n...@yahoo.com wrote: Looks nice to me! But no ivory towers around to pillage. (However planting a few seeds is almost always a good idea) Cheers, Alan From: Charles Perkins ch...@kuracali.com To: Fundamentals of New Computing fonc@vpri.org Sent: Friday, February 8, 2013 3:52 PM Subject: [fonc] yet another meta compiler compiler While we're all waiting for the next STEP report I thought I'd share something I've been working on, inspired by O'Meta and by the Meta-II paper and by the discussions on this list from November. I've written up the construction of a parser generator and compiler compiler here: https://github.com/charlesap/ibnf/blob/master/SyntaxChapter.pdf?raw=true The source code can be had here: https://github.com/charlesap/ibnf Don't be fooled by the footnotes and references, this is a piece of outsider literature. I am a barbarian come to pillage the ivory tower. Yarr. Chuck ___ fonc mailing list fonc@vpri.org http://vpri.org/mailman/listinfo/fonc ___ fonc mailing list fonc@vpri.org http://vpri.org/mailman/listinfo/fonc ___ fonc mailing list fonc@vpri.org http://vpri.org/mailman/listinfo/fonc
