Perl 6 - the future is here, just unevenly distributed

IRC log for #marpa, 2016-07-20

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

All times shown according to UTC.

Time Nick Message
00:59 ronsavage1 joined #marpa
01:19 JPGainsborough joined #marpa
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
03:27 idiosyncrat_ joined #marpa
04:19 jdurand Jeffrey, FYI I think you pointed me to right direction when talking about span.
04:20 jdurand In fact I see that in R2.xs you have a meta layer that is assigning the length in the input stream as a value to an earley set ID - I believe that I can use the user's output stack instead of that to propagate a void*value -; okay I have to think more about that, but it really seems that marpa_r_value_set() is how I can link forest tree to user's stack
04:58 idiosyncrat_ jdurand: if we are talking about a C language interface
04:58 idiosyncrat_ ...
04:58 idiosyncrat_ I would not use marpa_r_value_set -- that sets the value in the Earley set and I created it as (probably misguided) efficiency hack.
04:59 idiosyncrat_ Token values belong in the upper layers is my thinking at this point, and not in the Earley sets (Libmarpa layer)
05:00 idiosyncrat_ In fact, I'm considering eliminating marpa_r_value_set in future versions of Libmarpa -- once I move everything into the Kollos layer that belongs there, I plan to remove all those things from Libmarpa.\
05:01 idiosyncrat_ Your earlier quiestions made me think you were asking in the context of programming at the Perl level.
05:05 jdurand I see - nevertheless you want to keep the integer value that is in marpa_r_alternative() don't you ? Then you need to keep at least the method marpa_r_earley_set_value() don' you °
05:28 jdurand AFK working day will start -; cheers
05:41 ronsavage joined #marpa
06:18 idiosyncrat_ jdurand: no. my thought would be to eliminate token value entirely from Libmarpa -- Libmarpa does nothing with it, just stores it and passes it back, so that as far as token value is concerned, Libmarpa is simply an over-complex array implementation
06:20 idiosyncrat_ Token value can be stored in the higher layers, when it is desired.  This makes sense because only the higher layers know what the types and implementation of token values needs to be, and only they may actually be able to implement it.
06:21 idiosyncrat_ But for now Libmarpa will stay as it is.  Only once I get Kollos written, and its functionality fully settled, will I eliminate the features no longer needed in Libmarpa.
06:21 idiosyncrat_ Good night!
06:47 pczarn joined #marpa
09:58 ronsavage joined #marpa
12:22 rgrinberg joined #marpa
12:26 rgrinberg joined #marpa
14:11 idiosyncrat_ joined #marpa
16:00 rgrinberg joined #marpa
16:43 furelypunctional joined #marpa
16:43 idiosyncrat_ joined #marpa
17:30 rgrinberg joined #marpa
17:56 pczarn joined #marpa
18:45 pczarn A talk about Tango, literate programming in Rust. https://skillsmatter.com/skillscasts/8372-tango-literate-programming-in-rust
18:50 pczarn Probably relevant since you improve upon Knuth's WEB.
18:55 pczarn Later, the speaker decided to replace graphviz with some simpler, WYSIWYG tool.
19:26 rgrinberg joined #marpa
20:37 VsyachePuz_ joined #marpa
21:35 furelypunctional joined #marpa
23:13 ronsavage joined #marpa
23:39 rgrinberg joined #marpa

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