[00:59:40] *** Quits: koo6 (~sirdancea@79.98.72.195) (Ping timeout: 246 seconds)
[01:43:33] *** Quits: JuanDaugherty (~juan@98.4.124.117) (Quit: Hibernate, reboot, exeunt, etc.)
[07:25:32] *** Quits: proteus-guy (~proteus-g@2405:9800:b408:bc31:dddf:c1f7:934:3f41) (Remote host closed the connection)
[09:00:11] *** Joins: koo6 (~koom@79.98.72.195)
[09:16:36] *** Joins: proteusguy (~proteus-g@183.89.209.163)
[09:53:57] *** Quits: proteusguy (~proteus-g@183.89.209.163) (Read error: Connection reset by peer)
[09:54:53] *** Joins: proteusguy (~proteus-g@183.89.209.163)
[09:55:10] *** Quits: proteusguy (~proteus-g@183.89.209.163) (Read error: Connection reset by peer)
[09:56:38] *** Joins: proteusguy (~proteus-g@183.89.209.163)
[10:01:06] *** Quits: proteusguy (~proteus-g@183.89.209.163) (Read error: Connection reset by peer)
[10:13:34] *** Joins: proteusguy (~proteus-g@183.89.209.163)
[14:21:39] *** Quits: proteusguy (~proteus-g@183.89.209.163) (Remote host closed the connection)
[17:02:35] *** Joins: lin (c4d775a9@gateway/web/freenode/ip.196.215.117.169)
[17:04:16] <lin> hi - can anyone tell me what 'lin' does in e.g english resource grammar, paradigms.gf expression like ' lin N (some expression) ?
[17:05:11] <inariksit> hi!
[17:05:27] <lin> hi!
[17:05:46] <inariksit> so, when you write a lincat for some cat, you make it into some concrete record type
[17:05:56] <lin> yes
[17:06:02] <inariksit> like lincat N = { s : NForm => Str ; g : Gender } or whatever
[17:06:07] <lin> yes
[17:06:19] <inariksit> you could use the same concrete record for many different lincats
[17:06:30] <inariksit> like lincat Adv = {s : Str} ; lincat Conj = {s : Str} ;
[17:06:41] <lin> ok
[17:06:42] <inariksit> to you these are identical, but the compiler inserts a hidden field
[17:07:03] <inariksit> so actually adverb is {s : Str ; lock_Adv : {}} and same for Conj
[17:07:11] <inariksit> now, these are not meant for humans to manipulate
[17:07:17] <lin> ok
[17:07:25] *** Quits: koo6 (~koom@79.98.72.195) (Ping timeout: 246 seconds)
[17:07:33] <inariksit> so now, you may write an oper to make a noun
[17:07:39] <lin> yes
[17:07:48] <inariksit> oper mkN : Str -> { s : NForm => Str ; g : Gender} ;
[17:08:10] <lin> ok
[17:08:11] <inariksit> or probably you would give the clumsy { s : NForm => Str ; g : Gender} ; some other name
[17:08:18] <inariksit> like oper Noun : Type = { s : NForm => Str ; g : Gender} ;
[17:08:23] <inariksit> so oper mkN : Str -> Noun
[17:08:31] <lin> yes
[17:08:43] <inariksit> but the thing is that technically mkN doesn't make Ns, but it makes Nouns
[17:08:50] <lin> ok
[17:09:06] <lin> but wuth same structure
[17:09:11] <inariksit> so now you have in your lexicon something like  fun car_N : N ;
[17:09:12] <lin> as N's
[17:09:19] <inariksit> and in concrete lin car_N = mkN "car"
[17:09:36] <inariksit> but if your mkN makes Nouns, then mkN "car" will not work here
[17:09:43] <lin> ok
[17:10:09] <inariksit> instead of inserting lock_N field manually, you can just wrap it in lin N (mkN "car")
[17:10:20] <inariksit> but actually it's better to have the definition of mkN already do the wrapping
[17:10:35] <inariksit> so in practice what happesn is that mkN, an API function, calls some internal Noun-making function
[17:10:50] <inariksit> say you have oper regN : Str -> Noun ; and oper irregN : Str -> Gender -> Noun ;
[17:10:57] <inariksit> then you can call both of them from mkN
[17:11:06] <inariksit> oper mkN = overload  {
[17:11:27] <inariksit> mkN : Str -> N = \s -> lin N (regN s) ;
[17:11:42] <inariksit> mkN : Str -> Gender -> N = \s,g -> lin N (irregN s g) ;
[17:11:43] <inariksit> }
[17:12:00] <inariksit> and that's how you can just write mkN without worrying anything about lock fields!
[17:12:54] <lin> ok - i see - thank you so much !!!!!!!
[17:13:01] <inariksit> no worries :)
[17:13:20] <inariksit> are you by any chance coming to the summer school next week?
[17:13:52] <lin> no
[17:15:08] <lin> will be at south africa one
[17:15:17] <inariksit> cool! ^^
[17:15:42] <lin> :) bye - thanks again for such detailed help
[17:16:08] <inariksit> come anytime if you have more questions! this channel will probably be more alive next week too
[17:16:11] <inariksit> because of the summer school
[17:16:26] <lin> okay cool
[17:16:55] *** Parts: lin (c4d775a9@gateway/web/freenode/ip.196.215.117.169) ()
[18:50:02] *** Joins: koo5 (~koom@79.98.72.195)
[23:49:05] *** Quits: koo5 (~koom@79.98.72.195) (Ping timeout: 240 seconds)