[02:00:22] *** Quits: spectie (~fran@unaffiliated/spectie) (Remote host closed the connection) [02:02:32] *** Joins: spectie (~fran@2001:8b0:951:728:7a31:c1ff:fecd:969e) [02:02:32] *** Quits: spectie (~fran@2001:8b0:951:728:7a31:c1ff:fecd:969e) (Changing host) [02:02:32] *** Joins: spectie (~fran@unaffiliated/spectie) [02:54:47] *** Joins: jmvanel (~jmvanel@cpe-76-170-89-180.socal.res.rr.com) [05:04:48] *** Quits: jmvanel (~jmvanel@cpe-76-170-89-180.socal.res.rr.com) (Read error: Operation timed out) [05:43:57] Looks like cabal build langs=-Cat with a minus sign, will allow me to build without Catalan. [05:46:34] But cabal copy still fails with message, GrammarCat.gf does not exist, even though it is in lib/src/catalan. [06:56:37] oh, that sounds weird [06:57:00] does cabal install work anyway? [06:57:21] or the makefile in lib/src? [07:47:58] Checking out 3.5.12 from January and installing it, it installs. [08:00:57] ok [08:01:14] I'll try to check later today [08:02:03] I did changes in catalan 2 weeks ago, but it was just paradigms and structural [08:12:46] With that version installed, I tried installing master again with 'make install' and 'cabal install' and they both install now. [08:40:06] *** Joins: chru (~christina@2001:638:504:2099:39fd:16a0:d542:830d) [08:44:53] hi. does any of you know how the pipe in the gf shell is implemented? [08:45:32] it has an interesting behavior: when you have 2 languages loaded and do "p | l", then you get linearizations if the parsing succeeds in the first language but fails when the parsing fails in the first language (as there's no AST to linearize). so it seems that only the first parse result is piped, not all of them... [08:46:26] chru: I don't know [08:46:58] actually one of my students noticed... [08:47:01] probably krasimir or thomas know, but they're not on irc [08:47:24] I haven't noticed that either :P [08:48:16] yeah GF piping is not always as robust as you would expect [08:48:32] ok [08:48:33] drbean: I'm just running cabal install on the latest code [08:48:36] so far everything going fine [08:49:24] I only have one language, English. [08:49:40] yep everything went fine with the installation [08:51:48] I hacked lib/src/abstract/Verb.gf, lib/src/api/Constructors.gf and lib/src/english/VerbEng.gf. [08:52:57] But I don't need the code there and removed it. [09:02:49] chru: I looked at the code at https://bitbucket.org/chru/grammars. [09:03:29] i still plan to write a short doc that explains what's going on. not sure whether it's very clear from the code alone... [09:07:06] It's more motivational. That and svn://svn.molto-project.eu/wp4 give me goals to head to in the future. I want to investigate SPARQL in terms of that. [09:07:48] yes! i started a concrete syntax for SPARQL at the summerschool, but i never finished it... [09:08:13] anyway, the grammars there are as modular as application grammars can get ;) [09:16:53] drbean: do you think you give me a quick walk through your QA code at some point? i'd be interested to see how exactly it works (but it's always tedious to read other people's code in detail) [09:25:08] There isn't anything there which isn't in the Computational Semantics with Functional Programming book, I think. [09:30:24] There, Cats were identified by the phon string. I needed just one predicate for plural and singular Cats, so used an 'inflections' function to do that. But that function became unnecessary with GF. [09:32:09] ah. yes, i was wondering what you need it for if you have GF. [09:32:23] so the label function in Transfer determines the question type on the basis of the parse tree? then what is takeAnswer doing? [09:32:24] It is a vestige from porting to GF. [09:32:46] 'inflections' is a vestige, that is. [09:33:53] ah, i see, "answer" is doing the interpretation... only for yes/no questions? [09:34:25] Oh, that is not in CSFP, of course. I'm asking the ESL students to ask either a WH, YN, or Tag question. [09:35:12] chru: I saw Aarne and mentioned the pipe thing, he said he's not sure how it's implemented but recommended to use p -lang= "" if there are many languages [09:36:07] inariksit: ok, thanks. that's actually also what i told the student. [09:36:44] drbean: so what happens with WH questions? (and what is a Tag question?) [09:36:53] Ambiguous questions may have 2 different answers. I'm assuming the answer they want is Yes. [09:38:34] I used Exists and Forall in the book, I think to search for entities that satisfy the predicate. [09:39:00] ok [09:39:22] You know what a tag question with a tag at the end of the sentence is, don't you? [09:39:58] now i know :) [09:41:41] I used the StackParser to keep track of agreement. [09:42:30] Which was great for relative clauses too. I haven't tackled them with GF yet. [11:42:11] Does anyone know if there exists a query language (such as sql) as a concrete language for gf? [11:44:34] for SPARQL yes [11:44:46] it's in the MOLTO repo, svn://svn.molto-project.eu/wp4 [11:44:57] svn co svn://svn.molto-project.eu/wp4 , then go to the directory YAQL [11:45:39] and you see modules with Eng, Rus, Spa etc. and SPARQL :) [11:48:44] this thread might also be interesting: https://groups.google.com/forum/#!topic/gf-dev/YsGSZLJ-rNo [11:50:34] Nice! Thanks :) [13:33:18] *** Parts: chru (~christina@2001:638:504:2099:39fd:16a0:d542:830d) () [15:22:29] *** Joins: chru (~christina@2001:638:504:2099:5dc0:648b:281:de7e) [15:25:41] *** Quits: Eidel (~eidel@c83-249-247-253.bredband.comhem.se) (Ping timeout: 240 seconds) [15:27:43] *** Joins: Eidel (~eidel@c83-249-247-253.bredband.comhem.se) [16:39:19] *** Joins: jmvanel (~jmvanel@rrcs-173-196-156-164.west.biz.rr.com) [17:14:20] *** Quits: jmvanel (~jmvanel@rrcs-173-196-156-164.west.biz.rr.com) (Quit: Quitte) [17:44:58] *** Quits: chru (~christina@2001:638:504:2099:5dc0:648b:281:de7e) (Quit: Leaving.) [19:31:25] ha, second time I actually had to use the worst case constructor for nouns in finnish lexicon [19:31:30] first time was in 2011 [20:53:23] *** Joins: myoriel (~Myoriel@aftr-88-217-181-76.dynamic.mnet-online.de) [21:49:06] *** Joins: Eidel_ (~eidel@c83-249-247-253.bredband.comhem.se) [21:51:00] *** Quits: Eidel (~eidel@c83-249-247-253.bredband.comhem.se) (Read error: Operation timed out) [22:57:25] *** Quits: myoriel (~Myoriel@aftr-88-217-181-76.dynamic.mnet-online.de) (Ping timeout: 255 seconds)