[07:09:52] *** Joins: proteusguy (~proteusgu@cm-114-109-129-168.revip13.asianet.co.th) [07:43:16] *** Quits: proteusguy (~proteusgu@cm-114-109-129-168.revip13.asianet.co.th) (Ping timeout: 244 seconds) [08:11:33] *** Joins: proteusguy (~proteusgu@mx-ll-14.207.209-205.dynamic.3bb.co.th) [09:36:06] *** Quits: proteusguy (~proteusgu@mx-ll-14.207.209-205.dynamic.3bb.co.th) (Ping timeout: 272 seconds) [09:47:58] *** Joins: proteusguy (~proteusgu@mx-ll-14.207.209-205.dynamic.3bb.co.th) [11:23:38] *** Joins: bluemanblue (~manjaro-g@cgn85-194-8-119.bredband.comhem.se) [11:47:12] *** Quits: proteusguy (~proteusgu@mx-ll-14.207.209-205.dynamic.3bb.co.th) (Ping timeout: 272 seconds) [13:14:02] I'm getting the following error when I try to run [13:14:02] lincat [13:14:02] ClSlash = {n : NounPhrase ; v : Verb2 }; -- slash clause, e.g. "(that) I see" [13:14:02] lin SlashV2 np v2 = {n = np ; v = v2 }; [13:14:02] Why doesn't it recognize Verb2? [13:14:03] english/MiniExtEng.gf:7: [13:14:05] Happened in the renaming of ClSlash [13:14:07] constant not found: Verb2 [13:14:09] given MiniGrammarEng, MiniLexiconEng, MiniGrammarEng, MiniLangEng, [13:14:11] MiniExtEng [13:14:15] However, it is defined in MiniResEng. [13:14:17] [warrick@warrick-pc GF-Extension-Grammar]$ grep Verb2 english/* [13:14:19] english/MiniExtEng.gf: ClSlash = {n : NounPhrase ; v : Verb2 }; -- slash clause, e.g. "(that) I see" [13:14:21] english/MiniExtEng.gf: --Verb2 : Type = Verb ** {c : Str} ; [13:14:23] english/MiniGrammarEng.gf: V2 = Verb2 ; [13:14:25] english/MiniResEng.gf: Verb2 : Type = Verb ** {c : Str} ; [13:14:27] but MiniREsEng it is called by MiniGrammarEng [13:14:29] [warrick@warrick-pc GF-Extension-Grammar]$ grep MiniResEng english/* [13:14:31] Binary file english/MiniExtEng.gfo matches [13:14:33] english/MiniGrammarEng.gf:concrete MiniGrammarEng of MiniGrammar = open MiniResEng, Prelude in { [13:14:35] Additionally, I'm now keeping all my code on github. [13:14:37] https://github.com/wmacmil/GF-Extension-Grammar [13:18:32] why isn't MiniResEng included in "given ..." [13:21:54] If I change to [13:21:54] ClSlash = {n : NounPhrase ; v : V2 }; -- slash clause, e.g. "(that) I see" [13:21:54] I now get an alternative warning, (not error though) [13:21:54] MiniExtEng.gf: [13:21:54] MiniExtEng.gf:22: [13:21:55] Happened in linearization of SlashV2 [13:21:57] Warning: missing lock field lock_V2 [13:26:49] i'm guessing the first error is an inheritance issue? [13:51:58] i'm going for a second so if anyone does answer, please send to wmacmil@gmail.com [13:52:04] *** Quits: bluemanblue (~manjaro-g@cgn85-194-8-119.bredband.comhem.se) (Remote host closed the connection) [14:11:04] *** Joins: proteusguy (~proteusgu@184.82.200.177) [14:30:35] *** Joins: bluemanblue (~manjaro-g@static-88.131.7.244.addr.tdcsong.se) [14:37:25] cool, got it with [14:37:26] concrete MiniExtEng of MiniExt = MiniLangEng ** open MiniResEng, Prelude in { [14:45:14] bluemanblue: by the way, if you are away and someone has answered in the meantime you can check the backlog at http://www.grammaticalframework.org/irc/ [14:48:59] @daherb noted & thank you! [15:09:14] *** Quits: bluemanblue (~manjaro-g@static-88.131.7.244.addr.tdcsong.se) (Ping timeout: 272 seconds) [15:19:27] *** Joins: bluemanblue (~manjaro-g@static-88.131.7.244.addr.tdcsong.se) [16:05:04] Am wondering about how to address word order in QCl. I get an AST [16:05:04] UseQCl TSim PNeg (QuestVP who_IP (AdvVP (ComplV2 find_V2 (UsePron youSg_Pron)) now_Adv)) [16:05:04] which linearizes to [16:05:04] does who not find you now [16:05:04] The easiest fix would just be to have it return a Cl, which would then map to a MakeCl, however, this isn't consistent with our semantic categories. [16:05:05] so there is a feature of IP pronouns that reverse its order with the verb. Do you recommend adding a Boolean feature to all pronouns like isIP and then passing that to the UseQCl function which would then run a case analysis on isIP to determine if the evaluation is like in the original UseQCl or UseCl? [16:05:09] More generally, I find the iteration of types like this to address these kinds of issues kind of goes against the idea of making things as modular as possible. I know I'm still quite naive about GF, but the amount of work to ammend something simple like this seems disproportianate to the benefit. Has this been discussed much in the community? [16:13:19] *** Quits: bluemanblue (~manjaro-g@static-88.131.7.244.addr.tdcsong.se) (Ping timeout: 248 seconds) [16:39:00] *** Joins: bluemanblue (~manjaro-g@edu245202.eduroam.gu.se) [16:39:42] hi bluemanblue! I was just going out but I can have a quick comment [16:39:59] very welcome [16:40:33] as for your latest question, I don't understand what you mean with lack of modularity? I think it's just a feature; natural language is complex, and sometimes you need to have complex rules to model some phenomenon [16:41:11] if we had an abstract syntax only for English, it's possible that we could make a nicer design, but the point of the RGL abstract syntax is to be very general, and it's always a compromise [16:41:24] still a better deal for English than, say, Basque :-P [16:43:14] noted, and appreciated [16:45:14] it's a common design that intermediate categories are a collection of fields, and at later stages, say PredVP or UseCl, the word order is finally decided [16:45:29] it's also possible to have a table from e.g. word order to string [16:46:03] and to be prepared for all things that can be subjects, it sounds reasonable to have a boolean feature, if they behave differently [16:46:18] s/feature/field/ [16:46:56] in many languages it's a common thing to have isPron : Bool field in NPs, because there's UsePron : Pron -> NP, but pronouns often behave very differently [16:47:33] e.g. in romance languages they become clitics as objects, like "I see you" -> "je _te_ vois", not "*je vois toi" [16:48:48] the alternative would be to have a separate PredVP for NP -> VP -> Cl, and Pron -> VP -> Cl [16:49:35] okay that was it, now I'm off ^^ [16:50:23] thank you inariksit! [17:07:59] *** Quits: bluemanblue (~manjaro-g@edu245202.eduroam.gu.se) (Ping timeout: 244 seconds) [17:22:35] *** Quits: proteusguy (~proteusgu@184.82.200.177) (Ping timeout: 248 seconds) [18:46:40] *** Joins: proteusguy (~proteusgu@cm-58-10-155-156.revip7.asianet.co.th) [20:56:35] *** Joins: bluemanblue (~manjaro-g@2a00:801:3f1:eed2::2) [21:55:21] *** Quits: bluemanblue (~manjaro-g@2a00:801:3f1:eed2::2) (Ping timeout: 252 seconds)