Perl 6 - the future is here, just unevenly distributed

IRC log for #marpa, 2015-01-23

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

All times shown according to UTC.

Time Nick Message
02:13 ronsavage joined #marpa
03:00 jeffreykegler joined #marpa
05:16 ronsavage joined #marpa
06:35 basiliscos joined #marpa
06:44 rns joined #marpa
06:47 rns left #marpa
06:51 basiliscos1 joined #marpa
08:47 basiliscos joined #marpa
09:05 basiliscos joined #marpa
09:40 koo5 joined #marpa
13:36 koo5 joined #marpa
15:09 rns joined #marpa
15:12 rns jeffreykegler: just ran into that academia.edu thing -- http://homecookedtheory.com/arc​hives/2011/06/27/academia-edu/ -- https://medium.com/job-portraits/academia​-edu-is-reimagining-research-93e3ab1a3c95 -- thought it can be useful as a medium for the Marpa paper.
15:39 koo5 joined #marpa
16:30 jeffreykegler re http://irclog.perlgeek.de/​marpa/2015-01-23#i_9992199 -- Looks interesting.
16:31 jeffreykegler I've had a problem with similar services -- they require an academic affiliation, which I don't currently have.  Does academic.edu?
17:09 lwa joined #marpa
17:19 LLamaRider joined #marpa
21:32 jeffreykegler joined #marpa
21:51 ronsavage joined #marpa
21:54 jeffreykegler Re https://github.com/jeffreyk​egler/Marpa--R2/issues/233 -- github issue #233
21:56 jeffreykegler This describes an issue that's come up before -- if the ambiguous() method reports an ambiguity in a grammar, a subsequent value() method call fails with an error, because the ambiguous() method puts the parse tree into "forest" mode.
21:57 jeffreykegler There's a work around -- calling reset_series(0 resets the mode, so that value() will not work OK.
21:58 jeffreykegler This is not the smoothest interface possible but, until we move on from Marpa::R2, we are stuck with it ...
21:58 jeffreykegler however, I should document this (and thought I had, but apparently not).
21:58 jeffreykegler Anyway ...
21:59 jeffreykegler I think this issue, and the need for documentation, came up before.  Can anyone remind me of where?
21:59 jeffreykegler ronsavage: Do you remember this issue?
22:08 ronsavage jeffreykegler: Yes, I remember. Since I've never called reset_series() I must have done something else! I'll think about it. Actually, several things come to mind. I'll document them and let you know, probably via GG.
22:42 rns_ joined #marpa
23:02 ronsavage I've replied directly in the github issue #233.
23:36 flaviu joined #marpa
23:36 jeffreykegler ronsavage: [ back to IRC from the github issue ]
23:37 jeffreykegler Yes, your alternate approach to using Marpa is very interesting ...
23:37 jeffreykegler I've been presenting it as a syntax-driver parsing that allows procedural parsing, giving you the best of both worlds.
23:39 jeffreykegler And your approach is more to see Marpa as a powerful syntax-driven framework for procedural parsing.
23:40 jeffreykegler You should find some of the new features I'm putting into Marpa::R2 handy.
23:53 ronsavage New features! Drool, drool..........
23:53 jeffreykegler Yes, the discard events, but as side effects
23:54 jeffreykegler the ability to initialize all events either to 'off' or 'on' ...
23:55 jeffreykegler and the ability to access g1 spans -- g1 start location and length in g1 location terms.
23:55 jeffreykegler The g1 span information is motivated by the fact that, if you collect but discard information, and have a parse tree, g1 location is the only general way to sync them up.
23:56 jeffreykegler * but discard information -> both discard information

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