[00:59:36] *** Joins: Eidel_ (~eidel@c83-249-241-114.bredband.comhem.se) [01:03:04] *** Quits: Eidel (~eidel@c83-249-241-114.bredband.comhem.se) (Ping timeout: 260 seconds) [03:31:53] *** Joins: Masxmasx (~masx@2001:41d0:a:2340::1) [03:33:10] *** Quits: MasX (~masx@unaffiliated/masxmasx) (Ping timeout: 260 seconds) [04:05:47] *** Joins: Eidel (~eidel@c83-249-241-114.bredband.comhem.se) [04:08:52] *** Quits: Eidel_ (~eidel@c83-249-241-114.bredband.comhem.se) (Ping timeout: 250 seconds) [04:14:01] *** Joins: Eidel_ (~eidel@c83-249-241-114.bredband.comhem.se) [04:17:26] *** Quits: Eidel (~eidel@c83-249-241-114.bredband.comhem.se) (Ping timeout: 255 seconds) [06:12:35] *** Joins: drbean_ (~drbean@sac.nuu.edu.tw) [06:14:03] *** Quits: drbean (~drbean@113.196.171.151) (Ping timeout: 272 seconds) [07:07:05] *** Quits: crazydiamond (~crazydiam@dynamic-178-141-72-137.kirov.comstar-r.ru) (Quit: Ухожу я от вас (xchat 2.4.5 или старше)) [08:44:45] just putting this out there in case anyone is interested: [08:44:51] MT MOOC Starts 22nd October [08:44:57] We are pleased to announce a Massive Open On-line Course on Machine Translation that will start in early October. The course entitled: "Approaches to Machine Translation: rule-based, statistical and hybrid" is an opportunity to learn about the theory behind MT algorithms; how major on-line MT systems work; and how to manage open-source software to reproduce prototypes of such systems under a do-it-yourself basis. The MOOC is offered by the Universitat Polit [08:44:57] ècnica de Catalunya under the CANVAS platform. Students that successfully pass tests and assignments will be granted a certificate by the same University. [08:44:57] [08:44:58] For more information and registration, please visit: http://www.mt-mooc.upc.edu [09:26:40] a few apertiumers have registered :) [10:06:51] *** Joins: spectre (~fran@115.114.202.84.customer.cdi.no) [10:07:05] *** Quits: spectie (~fran@unaffiliated/spectie) (Read error: Connection reset by peer) [10:18:27] *** Quits: spectre (~fran@115.114.202.84.customer.cdi.no) (Ping timeout: 272 seconds) [10:19:34] *** Joins: spectre (~fran@115.114.202.84.customer.cdi.no) [11:52:11] *** Quits: spectre (~fran@115.114.202.84.customer.cdi.no) (Ping timeout: 272 seconds) [12:27:19] *** Joins: spectre (~fran@dhcp447-ans.wifi.uit.no) [12:40:26] *** Quits: Eidel_ (~eidel@c83-249-241-114.bredband.comhem.se) (Ping timeout: 240 seconds) [13:25:05] *** Joins: Eidel (~eidel@c83-249-241-114.bredband.comhem.se) [13:43:43] *** Quits: Eidel (~eidel@c83-249-241-114.bredband.comhem.se) (Ping timeout: 272 seconds) [13:50:08] *** Joins: Eidel (~eidel@c83-249-241-114.bredband.comhem.se) [14:05:30] *** Joins: Eidel_ (~eidel@c83-249-241-114.bredband.comhem.se) [14:08:30] *** Quits: Eidel (~eidel@c83-249-241-114.bredband.comhem.se) (Ping timeout: 246 seconds) [14:45:58] *** Quits: drbean_ (~drbean@sac.nuu.edu.tw) (Ping timeout: 276 seconds) [14:49:52] *** Joins: drbean (~drbean@124.219.83.123) [14:54:56] *** Quits: drbean (~drbean@124.219.83.123) (Ping timeout: 248 seconds) [15:06:00] *** Joins: drbean (~drbean@124.219.83.161) [15:12:37] *** Quits: drbean (~drbean@124.219.83.161) (Ping timeout: 276 seconds) [15:18:48] *** Joins: drbean (~drbean@124.219.82.176) [15:46:08] *** Quits: drbean (~drbean@124.219.82.176) (Ping timeout: 248 seconds) [15:47:13] *** Joins: drbean (~drbean@124.219.82.112) [16:14:26] *** Quits: spectre (~fran@dhcp447-ans.wifi.uit.no) (Ping timeout: 244 seconds) [17:57:04] *** Quits: drbean (~drbean@124.219.82.112) (Ping timeout: 276 seconds) [18:06:10] *** Joins: spectie (~fran@115.114.202.84.customer.cdi.no) [18:06:17] *** Quits: spectie (~fran@115.114.202.84.customer.cdi.no) (Changing host) [18:06:17] *** Joins: spectie (~fran@unaffiliated/spectie) [18:15:43] *** Joins: Eidel (~eidel@c83-249-241-114.bredband.comhem.se) [18:19:04] *** Quits: Eidel_ (~eidel@c83-249-241-114.bredband.comhem.se) (Ping timeout: 260 seconds) [21:31:25] I like the error messages :-P [21:31:27] expected: {s : ParamX.Number => ResEst.Case => Str; [21:31:27] isNum : Prelude.Bool; [21:31:27] n : ParamX.Number} -> {s : ResEst.NForm => Str} -> {s : ParamX.Number => ResEst.NPForm => Str; [21:31:30] a : ResEst.RAgr} [21:31:32] inferred: Type [21:32:41] hmm, i can't understand much of that :| [21:33:40] yeah :D [21:34:20] (this wasn't a "pls help" with very limited info, just thought it was funny :P) [21:34:27] :) [21:35:05] (but also maybe my brain just melted from being fooled to come to the programming logic seminar today and my sense of humour is seriously damaged) [21:36:00] :D [21:38:17] i bought a toilet brush today [21:38:20] they were on offer for 10kr [21:39:03] good story :D [21:41:55] hahah :D [21:42:09] don't worry [21:42:09] was that an equivalent experience to a proglog seminar [21:42:12] your sense of humour is fine! [21:42:43] also I swear I heard the phrase "type-directed type checking" during the talk, but now that I'm googling it (in quotes), I don't find any results [21:42:49] lol [21:42:56] type-undirected type checking [21:42:58] maybe the melting had happened already and I started replacing all words with "type" [21:43:52] haha [21:44:00] typey typey type type type [21:52:25] *** Joins: crazydiamond (~crazydiam@dynamic-178-141-72-137.kirov.comstar-r.ru) [21:55:50] grr I just want to rewrite the whole verb morphology [21:55:55] D: [21:56:01] of estonian ? [21:56:07] what is wrong with it ? [21:56:16] especially now that finnish has moved to the stem-based thing, there's no point keeping it similar [21:56:26] oh yeah [21:56:55] that thing :s [21:57:11] ok "whole verb morphology" is overstatement [21:58:10] more like the syntax part, there are some things that don't seem to be useful; there is VForm which is a morphological category and VIForm which is in VP [21:58:19] but I don't see a reason why we can't just use VForm everywhere [21:58:35] I'll ask aarne if he remembers some great reason :P [21:58:45] hmm [21:58:48] what is the 'I' ? [21:59:03] I don't even know [21:59:26] there are constructors VIFin, VIInf, VIPass and VIImper [21:59:32] so the I isn't infinitive [21:59:50] "indirect" ? :D [21:59:55] no idea [22:00:09] "interesting" "impenetrable" "invincible" [22:00:11] could be lots of things [22:00:20] ^^ [22:00:20] "invisible" :D [22:00:22] :D [22:00:23] haha [22:00:31] secret verb form [22:00:43] I wonder if anyone has actually done any applications in estonian using the grammar [22:01:07] trond is arriving in gothenburg and has problems compiling sme-smn :/ [22:01:12] well, in principle changing the insides shouln't break anything, because application grammarians should always use the api ^_^ [22:01:21] :D [22:01:27] there is a word for that! [22:01:30] in programming [22:01:31] abstraction! [22:01:46] :-O [22:02:11] also there's the downside that if I change things a lot, copypasting from finnish becomes harder [22:02:43] but then again, see the stem point [22:02:43] hmm, isn't it hard already ? [22:02:47] because of the stems stuff [22:02:51] yeah :) [22:03:00] also, is that going to be made default in new languages ? [22:03:10] haha maybe I should change the names of the finnish infinitive forms [22:03:18] hell with the Inf3Adess stuff [22:03:21] :D [22:03:22] yeah! [22:03:28] I don't know [22:03:36] you saw that it creates some problems in translation [22:03:45] yarmulke an experience and so on [22:03:49] and most languages don't need it [22:03:50] :D [22:03:52] maybe maltese would need it [22:03:58] well, also romance languages [22:04:03] cantándotela [22:04:19] though maltese has the non-concatenative thing, so maybe it's not so good idea [22:04:27] ah but that's just syntax [22:04:30] the pronominal suffixes are concatenative [22:04:36] magic &+ tokens to rescue [22:04:42] inariksit, but it makes a change inside the "stem" [22:04:51] can that be done with magic ? [22:04:57] honestly I don't even understand how does the finnish stemming work [22:05:00] ah yeah [22:05:22] I'd just add one more form to the verb [22:05:50] the point in finnish stemming is that instead of having a full inflection table of 8000000 forms [22:05:57] you have only 10 stems or so [22:06:00] aha [22:06:02] and then cases are added in syntax [22:06:17] so you'd just add the form for when you have certain sequences of pronouns following [22:06:51] for those forms that can have >=1) [22:07:07] yeah, you'd choose the CliticsAddedSpecialForm from the table [22:07:17] instead of PresPart or something [22:07:50] i guess you'd need it for inf/ger/imp [22:08:07] yeah [22:08:23] but it can be done without too much trouble [22:08:59] yeah [22:09:05] cool :) [22:09:14] changing the romance languages is a bit of a hassle though because of the functor [22:09:21] john could use something similar for maltese then [22:09:42] it's not enough just to change a lexical category, but all the functions that use them [22:09:48] there is a common romance functor ? [22:10:07] and if spanish is the only one that behaves that way, then we can just override the functions [22:10:10] yeah [22:10:20] hmm... sardinian has something similar i think [22:10:25] don't think italian does [22:10:27] french doesn't [22:10:35] portugues we don't have yet [22:10:40] portuguese is whack for other reasons [22:10:45] mesoclitics :s [22:10:50] what's that? [22:11:20] appears between the stem and tense affix [22:11:28] ooh [22:11:28] conquistar-se-á [22:11:32] cool [22:11:34] = "se conquistará" [22:14:14] yeah [22:14:32] :D [22:28:23] ok now I got more annoying error >__> [22:28:23] Internal error in GeneratePMCFG: [22:28:24] descend (CStr 340,CNil,CProj (LIdent (Id {rawId2utf8 = "fin"})) (CSel (App (App (QC (IC (Id{rawId2utf8 = "ResEst"}),IC (Id {rawId2utf8 = "Ag"}))) (QC (IC (Id {rawId2utf8 = "ParamX"}),IC (Id {rawId2utf8 = "Sg"})))) (QC (IC (Id {rawId2utf8 = "ParamX"}),IC (Id {rawId2utf8 = "P3"})))) (CSel (QC (IC (Id {rawId2utf8 = "ParamX"}),IC (Id {rawId2utf8 = "Pos"}))) (CSel (QC ((IC (Id {rawId2utf8 = "ParamX"}),IC (Id {rawId2utf8 = "Simul"}))) (CSel (App (QC (IC (Id {rawId2u [22:29:22] this one has menacing looking trees [22:29:34] yeah [22:30:15] the error message doesn't give line number or function or anything, but it must refer to this line [22:30:18] vps2str vps infform = (vps ! VIInf infform ! Simul ! Pos ! agrP3 Sg ).fin ; [22:30:28] https://optimisticpainter.files.wordpress.com/2010/09/forest_adjusted_blog.jpg [22:30:45] hmm [22:32:54] (replacing all rhs with a constant string works, so I just did that in order to get at least something testable :P) [22:33:03] (will whine to aarne tomorrow if I can't fix it :D) [22:33:11] :D [22:36:14] what is the constant string ? [22:38:22] guess [22:38:25] "hargle" [22:38:26] :D [22:38:44] harglebargle :D [22:38:47] :D [22:38:49] it was the only place I needed to do that [22:38:53] :)