Perl 6 - the future is here, just unevenly distributed

IRC log for #marpa, 2015-09-23

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

All times shown according to UTC.

Time Nick Message
01:48 ilbot3 joined #marpa
01:48 Topic for #marpa is now Start here: http://savage.net.au/Marpa.html - Pastebin: http://scsys.co.uk:8002/marpa - Jeffrey's Marpa site: http://jeffreykegler.github.io/Marpa-web-site/ - IRC log: http://irclog.perlgeek.de/marpa/today
02:52 ronsavage joined #marpa
03:48 ceridwen idiosyncrat, Is there a description of how you implemented the forest for Marpa up anywhere?  I looked in your theory paper and didn't see anything with a quick skim/search (I've read it before, and I didn't remember anything, but...).
03:49 ronsavage joined #marpa
03:56 Idiosyncrat joined #marpa
03:57 Idiosyncrat_ ceridwen: there is Scott's paper -- Libmarpa's ideas and Scott's are so similar, that it's not really worth a Libmarpa paper on the topic.
04:00 ceridwen Which paper are you thinking of?  Scott's paper covers the abstract theory, but I'm wondering about the concrete data structures.  Is there another paper I'm not aware of?
04:53 Idiosyncrat_ ceridwen: For the very gritty details of implementation, there's the Libmarpa code
04:57 Idiosyncrat_ My recollection was that Scott in fact got pretty nitty-gritty
04:58 Idiosyncrat_ she does leave choices like linked list vs. AVL vs. hash up to the reader
04:58 Idiosyncrat_ my most careful reading was my first, when I was still thinking the bocage was my invention :-)
04:59 Idiosyncrat_ and I recall being impressed with how her paper covered just about everything you'd want in a writeup of Libmarpa's bocages. :-)
05:02 Idiosyncrat_ I have to call it a night now.
05:02 Idiosyncrat_ ceridwen: I hope I answered your question.  If not, I'll backlog in the morning.
05:03 Idiosyncrat_ Good night!
06:33 ronsavage joined #marpa
07:26 mvuets joined #marpa
07:26 mvuets Hi! Just a quick heads-up. I've published all content of my "Processing toki pona with Perl" talk that I gave at Swiss Perl Workshop 2015. Check it out here - https://github.com/mvuets/talk-​processing-toki-pona-with-perl Bye! (-:
08:29 ronsavage joined #marpa
08:46 ronsavage joined #marpa
10:15 lwa joined #marpa
13:52 djns joined #marpa
18:40 idiosyncrat joined #marpa
19:01 MerittClarksdale joined #marpa
20:06 idiosyncrat I am toying with the idea of doing some self-made videos
20:07 idiosyncrat No big deals, but just to give Marpa a "face".
20:07 idiosyncrat Does anybody recommend a make for low-cost digital cameras.
20:08 idiosyncrat There's one on my laptop, of course, but it gives kind of a "fish eye" look to things, and the object is to make me less strange, not more. :-)
20:09 djns are you looking for one that will work on Linux?
20:12 djns I have a logitech V-U0026
20:13 djns also known as C170
20:20 idiosyncrat Yes, I should have mentioned.  I run Debian stable on both laptop and desktop.
21:30 ceridwen idiosyncrat, She talks about which operations have to be constant-time and the like, but doesn't discuss implementation choices, which is where I am now.  (I'm still trying to digest the discussion about hash tables not being constant time and decide whether I need to change my implementation of the SPPF.)
23:18 idiosyncrat ceridwen: My advice -- don't worry about constant time.
23:18 idiosyncrat It's important to theoreticians.
23:19 idiosyncrat I'm about the only actual implementer I know that insists on hand-crafting a O(n) alternative where a nice easy O(n log n) alterative is there to just grab.
23:20 idiosyncrat And I did so early in the development of Marpa.
23:20 idiosyncrat One could make the criticism that I should have invested my time elsewhere.
23:21 idiosyncrat You might want to note in the code where you are using an quasi-linear implementation, but claim linear for theory purposes.
23:21 idiosyncrat Even in Libmarpa, there are a few places where I did that, and left notes to that effect in the comments.
23:25 rns joined #marpa
23:33 rns Aria, idiosyncrat: how about adding, for completeness sake, Earley's An Efficient Context-Free Parsing Algorithm to Aria's list of parsing papers?
23:34 rns The thing is I've found a very good (academic-grade) translation of Earley's paper into Russian, which can also be added to Aria's list.
23:38 rns Tricky CSV parsing -- hard with Text::CSV, easy with Marpa::R2 -- http://stackoverflow.com/questions/3274​8557/parse-csv-file-with-line-breaks-an​d-additional-commas/32751046#32751046

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