Perl 6 - the future is here, just unevenly distributed

IRC log for #marpa, 2016-01-03

| Channels | #marpa index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
02:16 ronsavage left #marpa
02:18 ronsavage joined #marpa
02:24 ronsavage I released https://metacpan.org/release/Tree-Cladogram so now I'm thinking: What's next. Either it'll be a parser for regexps, or perhaps more likely, a parser for Image::Magick's command line utilities. The docs for that are here: http://imagemagick.org/scri​pt/command-line-options.php. We'll see.
06:23 kaare_ joined #marpa
11:15 koo7 joined #marpa
11:22 jdurand joined #marpa
11:23 jdurand I recently hitted the following exception in grammar's parse() method: "value() called when recognizer is not in tree mode. The current mode is "forest""
11:24 jdurand I understand I am not in the current "mode", though this certainly indicate something wrong in the grammar - how may I find what is going wrong...?
11:24 jdurand (I am not in the "tree" mode)
11:58 pczarn joined #marpa
13:29 Idiosyncrat joined #marpa
14:30 koo7 joined #marpa
14:50 jdurand Got it - this is my Marpa speedup routines in which Marpa::R2::Internal::Recognizer::TREE_MODE internal variable was not handled
16:18 Idiosyncrat joined #marpa
17:10 koo7 joined #marpa
17:15 Idiosyncrat On twitter, one commenter said that Marpa is "documented like a Ponzi scheme".
17:16 Idiosyncrat Not sure that's good, but she did keep reading. :-)
17:18 lucs Idiosyncrat: Yeah, I was wondering what that meant...
18:11 VsyachePuz Idiosyncrat: I understood it as "have a lot of marketing blaf and not enough in-depth explainations"
18:11 Idiosyncrat I'm hearing that a lot lately, and it perplexes me ...
18:12 Idiosyncrat there *is* a theory paper.
18:12 Idiosyncrat I mean, that is the ultimate and traditional form for in-depth explanations.
18:13 VsyachePuz paper is an obsolete form. Hyperlinked body of materials is what we expect
18:13 Idiosyncrat Hmmm ...
18:13 Idiosyncrat Like for example?
18:14 Idiosyncrat That is, can you give me an example of a very math-y project that has such documentation?
18:15 Idiosyncrat In terms of other parsers, Perl 6 grammars have just a few lines, and currently they're asking me how it behaves.
18:15 Idiosyncrat ANLTR does papers
18:15 Idiosyncrat PEG has papers
18:20 Idiosyncrat I wonder if the problem is that it's not immediately evident from Ron's site that there's a theory paper.
18:20 VsyachePuz new generation of learning tools should allow interactivity (issue tracking at least). One wxample is Wiki tutorials
18:21 Idiosyncrat Perhaps, but nobody doing anything as math-y as I am doing is pioneering this, and I'm overloaded as it is.
18:21 VsyachePuz https://en.wikipedia.org/wiki/Automata_theory
18:23 Idiosyncrat OK, I see.
18:23 VsyachePuz The idea is that the amount of text has much more volume than in a single paper, but reader is free to choose directions
18:23 Idiosyncrat But there still is the matter of time for someone to create it.
18:24 Idiosyncrat The people doing the original work in automata theory are writing papers, I'd imagine.
18:24 Idiosyncrat With Marpa, I'm already doing something unprecedented -- the inventors of parsing algorithms have not in the past implemented them.
18:25 Idiosyncrat For yacc, Knuth did some theory and left it to ...
18:25 Idiosyncrat DeRemer who did some more theory and left it to ...
18:26 Idiosyncrat Ullmann who researched it and thought it out and advised ...
18:26 Idiosyncrat Steve Johnson who actually wrote yacc.
18:26 VsyachePuz so, the problem is in economics, right? You can't build set of agreements which will allow you to organize collective work
18:27 Idiosyncrat Err, that's one way of structuring the issue.  Does it move things forward in any way?
18:28 VsyachePuz DARPA goes this way, ant yes that creates an internet
18:29 Idiosyncrat DARPA had beaucoup bucks -- I've been self-funded since 2007 and I ain't rich.
18:29 Idiosyncrat Give me the bucks they gave DARPA and you can expect *a lot* more to happen.
18:30 VsyachePuz other people (bill gates for example) have he same amount of heads, hands and legs as you. But they have more money. So you strategy can be better
18:31 Idiosyncrat Btw, I think Loup-Vaillant's tutorials may be steps in the direction you're talking about ...
18:31 Idiosyncrat and your comments would explain why they were so well-received.
18:31 VsyachePuz http://loup-vaillant.fr/tutorials/earley-parsing/
18:32 Idiosyncrat Right.
18:32 Idiosyncrat Loup-Vaillant hasn't got as far as Leo yet, something I hope the Marpa book (in progress) will resolve.
20:13 koo7 joined #marpa
20:47 VsyachePuz exactly my case - "Elizabeth Scott described a method that works, but I totally failed to comprehend it."
20:47 VsyachePuz http://loup-vaillant.fr/tuto​rials/earley-parsing/parser
21:28 ronsavage joined #marpa
21:29 VsyachePuz i completed the first read. Good tutorial because it uses good words (which comes from other topics, learned earlier). But I still don't understood how SPPF works. May be now i need to reread that paper. Now I have more motivation to look at C# earley parser which was still not packaged by me - https://github.com/coonsta/earley

| Channels | #marpa index | Today | | Search | Google Search | Plain-Text | summary