Perl 6 - the future is here, just unevenly distributed

IRC log for #marpa, 2015-08-06

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

All times shown according to UTC.

Time Nick Message
00:55 idiosyncrat 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
02:55 koo7 joined #marpa
02:55 koo7 left #marpa
04:20 ronsavage joined #marpa
06:14 rns joined #marpa
06:22 ronsavage joined #marpa
08:09 ronsavage joined #marpa
10:08 purmou joined #marpa
12:29 koo7 joined #marpa
12:36 aredridel joined #marpa
12:40 koo7 if it might interest anyone, ive built a parser for a subset of n3. The gotacha and at the same time the interesting thing is that its currently implemented as a builtin of our reasoner, over libmarpa. Grammar is in the form of linked data, loaded by a basic nquads parser, then a generic parse tree is created, then rules for the reasoner are created from that
13:01 LLamaRider joined #marpa
13:29 purmou joined #marpa
14:28 djns joined #marpa
17:03 idiosyncrat joined #marpa
17:31 idiosyncrat This is the project at https://github.com/naturalog/tauchain ?
17:34 idiosyncrat I looked at it and downloaded Ohad Asor's paper -- http://tauchain.org/tauchain.pdf
17:54 lwa joined #marpa
17:57 koo7 right
17:57 koo7 paper isnt too bad, code still has long way to go
17:59 koo7 the marpa bit is currently broken as im doing some refactoring
18:06 idiosyncrat koo7: I noticed a new notation in the Github listing for the libmarpa entry
18:07 idiosyncrat Do you have it in there as a submodule?  How did you do that?
18:11 koo7 git submodule add
18:26 idiosyncrat Let us know how it works for you.  IIRC there's two other ways to accomplish the same thing, all of which looked dicey to me in one way or another.
18:27 idiosyncrat Currently, I just copy code over -- this has its problems, but none of the problems are surprises.
18:30 koo7 i never had a single problem with this approach
18:31 koo7 although my use cases are small repos, personal projects
18:33 koo7 git can be pretty surprising tho, until really mastered
18:33 idiosyncrat By the way, for anyone who does the "copy over" approach, I do a "git log -n 5", so that I can be sure which commit you've copied.
18:33 idiosyncrat s/you've/I've/
18:34 idiosyncrat In my case, it was for Kollos, which will do a severe fork of the Libmarpa code anyway, so it makes sense to break the ties to the old code.
18:35 koo7 ideally by proper forking:)
18:36 idiosyncrat By "proper", do you mean a Github fork?
18:36 koo7 yeah
18:37 idiosyncrat Well, the Kollos repo is already well along, and it's not a Libmarpa fork.
18:39 koo7 yeah, would only make sense if the fork was included in the kollos repo as a submodule
18:41 koo7 actually, i was wondering how to make building our project with libmarpa easier, i guess better than submoduling the whole libmarpa, i should build a "dist" for others
18:42 koo7 ive also wondered if it would be possible to link against the Marpa R2 lib that is in ubuntu
18:42 koo7 ran into perl deps ofc, which i didnt manage to satisfy
20:00 idiosyncrat Actually, a Libmarpa dist would be useful -- I've never created a real distribution discipline for Libmarpa, of the sort which exists for Marpa::R2
20:01 idiosyncrat By discipline, I mean a careful statement of which commit is which release, and what kind of release it is.
21:16 pczarn joined #marpa
21:22 koo7 joined #marpa
21:24 pczarn Is there some reason for keeping completed items? I noticed that previous Earley sets are accessed exclusively through indexes by postdot symbol
22:18 pczarn joined #marpa
22:18 pczarn I guess they may be needed for progress reports
22:32 ronsavage joined #marpa
23:02 ronsavage Haxe and Typescript both compile-to-JS: http://blog.onthewings.net/20​15/08/05/typescript-vs-haxe/
23:06 ronsavage A trenchant defence of Javascript: http://www.crockford.com/j​avascript/javascript.html
23:08 idiosyncrat joined #marpa
23:17 idiosyncrat joined #marpa
23:18 idiosyncrat pczarn: As has been pointed out in the literature, completed items can be deduced from the others, and might be considered a memoization.
23:18 idiosyncrat But it's a useful one.
23:19 idiosyncrat koo7: (and others) Re libmarpa distributions, I forgot something important -- there *is* an experimental level Debian distribution of libmarpa
23:28 koo5 ah, good

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