[00:06:45] *** Joins: fotonzade (~fotonzade@78.189.158.97) [00:43:37] *** Quits: fotonzade (~fotonzade@78.189.158.97) (Ping timeout: 276 seconds) [01:53:24] *** Joins: fotonzade (~fotonzade@78.189.158.97) [01:55:38] *** Quits: fotonzade (~fotonzade@78.189.158.97) (Client Quit) [08:21:18] *** Joins: venicecat (caa61422@gateway/web/freenode/ip.202.166.20.34) [08:29:56] hmm what am i doing wrong with want_VV of type VV? [08:30:37] I am puzzling over the long error message "Happened in linearization of want_VV no overload instance of mkVP for {s : VVForm => Str; lock_VV : {}; p : Str;..." [08:30:56] i wrote want_VV : Deontic in abstract syntax with Deontic = VV in the concrete syntax. [08:31:18] The lin rule is just as RGL described - want_VV = mkVP want_VV (mkVP withdraw_V) [10:19:19] mkVP builds a VP, not a VV [10:20:23] if you want something to include want and witjdraw, thats a VP built out of two components, wantVV and a VP with withdraw [10:20:30] sorry typos, on mobile [10:22:00] if that doesnt help, can you paste some more of the grammar? what other stuff do you have in Deontic, should they be just verbs like want, should, must, and what other functions use Deontic? [11:15:27] *** Quits: sirdancealot (~sirdancea@79.98.72.195) (Ping timeout: 240 seconds) [11:43:42] *** Joins: fotonzade (~fotonzade@78.189.158.97) [11:47:29] thanks inariksit! [11:49:09] right now, deontic just consists of want; [11:50:09] i am working through the resource library hoping to git gud at it.. [11:51:26] and the modular system too [11:51:46] i have pasted the code into gist https://gist.github.com/anonymous/23068853fc7e253a04824e28a99f9313 [11:51:55] looking forward to your advice! [12:07:31] okay, so you should just linearise Want as want_VV [12:07:51] and move the mkVP into Pred [12:08:15] Pred party deontic action = mkUtt (mkCl party (mkVP deontic action)) ; [12:08:53] *** Quits: fotonzade (~fotonzade@78.189.158.97) (Quit: Leaving) [12:09:33] actually no, sorry, you can just use mkCl NP -> VV -> VP -> Cl [12:09:41] instance of mkCl directly, so [12:10:11] Pred party deontic action = mkUtt (mkCl party deontic action) ; [12:26:16] thanks, it works! [12:26:20] woo! [12:27:03] so even though RGL says using this combination (mkVP want_VV (mkVP sleep_V)) [12:27:39] i should just leave out the mkVP and linearise want directly [12:55:25] well it doesn't really matter, the API offers many instances of how to form a Cl, a VP etc, I just offered the NP -> VV -> VP -> Cl instance because it saves you some typing [12:59:34] *** Joins: fotonzade (~fotonzade@78.189.158.97) [13:09:49] *** Quits: venicecat (caa61422@gateway/web/freenode/ip.202.166.20.34) (Ping timeout: 260 seconds) [16:55:23] *** Quits: fotonzade (~fotonzade@78.189.158.97) (Ping timeout: 268 seconds) [20:19:10] *** Joins: sirdancealot (~sirdancea@79.98.72.195) [23:02:39] *** Joins: fotonzade (~fotonzade@78.189.158.97) [23:10:04] *** Quits: fotonzade (~fotonzade@78.189.158.97) (Quit: Leaving) [23:10:26] *** Quits: proteusguy (~proteus-g@2605:6000:151a:801c:245b:1c0d:5aea:5eba) (Quit: Leaving)