[02:01:57] <freeside> slack is down
[02:02:00] <freeside> hi all
[02:02:11] <inariksit> hi!
[02:02:40] *** Joins: venicecat (b0475f00@gateway/web/freenode/ip.176.71.95.0)
[02:02:48] <vin-ivar> s/down/awful
[02:03:00] <inariksit> hi everyone
[02:03:04] <inariksit> feels so cosy :3
[02:03:27] <venicecat> \o/
[02:03:32] <vin-ivar> yeah there's three whole people around ;D
[02:03:35] <freeside> mysig
[02:04:08] *** Joins: wabisabiyo (b0475f00@gateway/web/freenode/ip.176.71.95.0)
[02:04:15] <inariksit> hello!
[02:04:23] <wabisabiyo> hello!
[02:06:59] <freeside> hello!
[02:07:54] <inariksit> vin-ivar: kristian is suggesting his flat in estonia
[02:08:30] <vin-ivar> ooh
[02:08:34] <vin-ivar> estonia sounds gr8 :D
[02:08:38] <inariksit> ^_^
[02:08:42] <vin-ivar> gr9, even
[02:10:40] <freeside> how about a barn in rural Latvia
[02:10:57] <freeside> they have chickens, which are Kyckling in Swedish
[02:11:00] <inariksit> that would probably work too for kristian
[02:12:30] *** Quits: wabisabiyo (b0475f00@gateway/web/freenode/ip.176.71.95.0) (Quit: Page closed)
[02:34:56] *** Quits: koom_ (~koo5@109.107.211.244) (Ping timeout: 260 seconds)
[02:51:04] *** Quits: venicecat (b0475f00@gateway/web/freenode/ip.176.71.95.0) (Ping timeout: 260 seconds)
[05:31:46] *** Quits: proteusguy (~proteus-g@2405:9800:bc10:1ca:406c:6d42:a278:38df) (Remote host closed the connection)
[07:00:32] *** Joins: proteusguy (~proteus-g@180.183.98.27)
[09:29:02] *** Joins: koom (~koo5@79.98.72.195)
[09:59:27] *** Quits: proteusguy (~proteus-g@180.183.98.27) (Ping timeout: 240 seconds)
[10:39:10] *** Joins: proteusguy (~proteus-g@180.183.14.107)
[10:43:08] *** Quits: proteusguy (~proteus-g@180.183.14.107) (Excess Flood)
[10:44:08] *** Joins: proteusguy (~proteus-g@180.183.14.107)
[13:56:37] *** Quits: proteusguy (~proteus-g@180.183.14.107) (Remote host closed the connection)
[14:15:13] *** Joins: proteusguy (~proteus-g@cm-27-145-90-73.revip12.asianet.co.th)
[14:21:06] *** Quits: proteusguy (~proteus-g@cm-27-145-90-73.revip12.asianet.co.th) (Max SendQ exceeded)
[14:22:02] *** Joins: proteusguy (~proteus-g@cm-27-145-90-73.revip12.asianet.co.th)
[15:00:06] *** Joins: e3928a3bc (~user@2804:14d:5cd8:8447::4)
[15:44:02] *** Joins: venicecat (c024e1f8@gateway/web/freenode/ip.192.36.225.248)
[15:57:23] *** Quits: proteusguy (~proteus-g@cm-27-145-90-73.revip12.asianet.co.th) (Remote host closed the connection)
[16:02:46] <e3928a3bc> hey people
[16:03:02] <e3928a3bc> I have a doubt concerning an exercise from the book (again!)
[16:03:35] <e3928a3bc> it's exercise 5 from chapter 3
[16:04:03] <e3928a3bc> concrete E5conc of E5 = {
[16:04:03] <e3928a3bc>   lincat
[16:04:03] <e3928a3bc>     S = {a,b,c : Str} ;
[16:04:03] <e3928a3bc>     FS = Str ;
[16:04:06] <e3928a3bc>   lin
[16:04:09] <e3928a3bc>     Empty = {a,b,c = []} ;
[16:04:12] <e3928a3bc>     Iter s = {a = "a" ++ s.a; b = "b" ++ s.b ; c = "c" ++ s.c} ;
[16:04:16] <e3928a3bc>     Print s = s.a ++ s.b ++ s.c ;
[16:04:18] <e3928a3bc> }
[16:04:30] <e3928a3bc> this is my concrete grammar, but the results are off, and I can't find out why
[16:05:08] <e3928a3bc> it outputs strings such as "a a a a" instead of "a a b b c c"
[16:15:41] *** Joins: proteusguy (~proteus-g@2405:9800:bc10:1ca:e544:2d8d:f77e:e2ef)
[16:15:59] <vin-ivar> what are you trying to linearise?
[16:16:39] *** Quits: venicecat (c024e1f8@gateway/web/freenode/ip.192.36.225.248) (Ping timeout: 260 seconds)
[16:56:19] <inariksit> vin-ivar: it's the language aⁿbⁿcⁿ
[16:58:13] <vin-ivar> no i mean
[16:58:17] <vin-ivar> what are you typing in :D
[17:00:10] <inariksit> ah right
[17:19:22] <e3928a3bc> I'm typing ">  gr | l" at the prompt
[17:21:03] <e3928a3bc> it will parse "a a" no problem, but not "a a b b c c"
[17:21:17] <e3928a3bc> and I have no idea why..
[17:24:20] <daherb> e3928a3bc: what is your startcat? gf uses by default S i think. so you can try 'p -cat=FS "a b c"'
[17:25:39] <vin-ivar> ^
[17:26:04] *** Quits: vin-ivar (~vin-ivar@unaffiliated/vin-ivar) (Quit: WeeChat 1.0.1)
[17:26:28] *** Joins: vin-ivar (~vin-ivar@unaffiliated/vin-ivar)
[17:28:04] <e3928a3bc> it is FS! my abstract grammar: abstract E5 = {
[17:28:05] <e3928a3bc>   flags
[17:28:05] <e3928a3bc>     lexer=concat ; unlexer = concat ; startcat = FS ;
[17:28:05] <e3928a3bc>   cat
[17:28:08] <e3928a3bc>     S ; FS ;
[17:28:11] <e3928a3bc>   fun
[17:28:14] <e3928a3bc>     Print : S -> FS ;
[17:28:17] <e3928a3bc>     Iter : S -> S ;
[17:28:21] <e3928a3bc>     Empty : S ;
[17:28:24] <e3928a3bc> }
[17:28:27] <e3928a3bc>  
[17:28:29] <inariksit> yeah so that is the problem, just try 'p -cat=FS "a b c"'
[17:28:31] <e3928a3bc> (I don't know how to setup the flags correctly, but that doesn't seem to be the problem, I hope)
[17:28:53] <inariksit> I think the lexer and unlexer flags are deprecated?
[17:29:24] <e3928a3bc> hum, ok. it parses correctly using 'p -cat=FS "a b c"'!
[17:29:29] <e3928a3bc> but why?
[17:29:30] <inariksit> but anyway, the point is that if you try to parse your string in the S category, and it doesn't have a field called s, weird stuff happens
[17:29:42] <inariksit> so there are two weird things that are probably not well-documented
[17:29:52] <inariksit> 1) S is by default a startcat
[17:29:56] <inariksit> unless you specify another
[17:29:57] <e3928a3bc> shouldn't the 'startcat=FS' take care of this?
[17:29:59] <inariksit> yes
[17:30:06] <inariksit> ah right
[17:30:26] <e3928a3bc> is 'S' overriding my setup?
[17:31:19] <inariksit> 2) if you have a lincat with multiple fields, and the string that you try to parse is spread over different fields, it won't work
[17:31:56] <inariksit> sorry I haven't slept properly and my brain isn't working now; did you have already fhe 'flags startcat=FS' when you first reported the problem?
[17:32:17] <inariksit> if you specified the startcat, then S shouldn't override anything
[17:36:45] <e3928a3bc> not a problem! yes, I did have the 'flags startcat=FS' beforehand!
[17:37:48] <inariksit> okay
[17:37:50] <inariksit> this works for me:
[17:37:52] <inariksit> concrete E5conc of E5 = {
[17:37:52] <inariksit> lincat
[17:37:52] <inariksit>      S = {a,b,c : Str} ;
[17:37:52] <inariksit>      FS = { s : Str } ;
[17:37:52] <inariksit>    lin
[17:37:53] <inariksit>      Empty = {a,b,c = []} ;
[17:37:53] <inariksit>      Iter s = {a = "a" ++ s.a; b = "b" ++ s.b ; c = "c" ++ s.c} ;
[17:37:54] <inariksit>      Print s = { s = s.a ++ s.b ++ s.c } ;
[17:38:06] <inariksit> so exactly like before, but FS is linearised as a {s : Str} instead of Str
[17:39:52] *** Joins: e3928a3bc2 (~yaaic@2804:14d:5cd8:8447:89d8:e1e2:f7c0:518b)
[17:40:38] <inariksit> did you see my latest comment? if you change lincat FS = { s : Str } ;, and  Print s = { s = s.a ++ s.b ++ s.c } ; it works for me
[17:41:20] <e3928a3bc> yes! I was setting up IRC on mobile so that I can still chat
[17:41:42] <e3928a3bc> I'll try it, but hum... do you know why it wouldn't work before?
[17:41:47] <inariksit> no idea
[17:41:56] <inariksit> just a general "there's something special in the field called s" thing
[17:42:24] <e3928a3bc> ok! maybe one day I'll get to read the source, but I must learn how to use gf first :)
[17:43:06] <e3928a3bc> I tried to use '+' before, so that I wouldn't have spaces between the characters, but I can't use it with runtime variables
[17:43:14] <inariksit> yeah exactly, all tokens must be known at runtime
[17:43:22] <inariksit> so you can use + in functions that create tokens
[17:43:27] <inariksit> s/functions/opers/
[17:43:39] <e3928a3bc> my next attempt was trying to use the concat lexer to remove the spaces, but as you've said, they are now deprecated?
[17:43:58] <e3928a3bc> that is, I should use them from the terminal, and not include the in the grammar?
[17:43:59] <inariksit> so e.g. when constructing an inflection table with a helper oper, and then at runtime there is the full inflection table within the noun/verb/whatever
[17:44:14] <inariksit> ah, for glueing things without spaces at runtime, we have a new shiny thing
[17:44:19] <inariksit> the BIND token
[17:44:36] <inariksit> if you're interested in details, http://www.aclweb.org/anthology/W15-3305
[17:44:48] <inariksit> if you just want to use it, there's a token called BIND in Prelude
[17:45:38] <inariksit> and even functions called glue, glueOpt, noglueOpt https://github.com/GrammaticalFramework/GF/blob/master/lib/src/prelude/Prelude.gf#L102-L105
[17:45:56] <inariksit> so you can say e.g.
[17:46:25] <inariksit> Iter s = {a = glue "a" s.a; b = glue "b" s.b ; c = glue "c" s.c} ;
[17:47:13] <inariksit> and then you have to open Prelude, so just begin your concrete syntax with the following:
[17:47:16] <inariksit>  
[17:47:17] <inariksit> concrete E5conc of E5 = open Prelude in {
[17:51:28] <e3928a3bc> I'm moving to the mobile, will answer soon. thanks a lot, inariksit !
[17:51:35] *** Quits: e3928a3bc (~user@2804:14d:5cd8:8447::4) (Quit: ERC (IRC client for Emacs 25.2.1))
[17:53:47] <inariksit> so the BIND token: this is what happens with the version where Iter uses glue
[17:53:47] <inariksit> E5> l Print (Iter Empty)
[17:53:48] <inariksit> a &+ b &+ c &+
[17:54:09] <inariksit> but just give it the argument -bind, and this happens:
[17:54:11] <inariksit>  
[17:54:11] <inariksit> E5> l -bind Print (Iter Empty)
[17:54:12] <inariksit> abc
[17:54:14] <inariksit>  
[17:54:36] <inariksit> for parsing in the standard GF shell, you need to input the &+ characters too, which is kinda clumsy
[17:55:09] <inariksit> we have at the moment 2 runtimes, one in Haskell and one in C, and the C one supports also parsing just "abc" and not "a &+ b &+ c"
[17:55:30] <inariksit> but the one used in the GF shell is the older Haskell runtime, which doesn't support just parsing "abc"
[17:56:29] <inariksit> if you use a compiled GF grammar from another program, and you use the BIND token, then it makes sense to use the C runtime. we have bindings to several programming languages, and the API is almost identical to the old Haskell runtime
[17:57:14] <inariksit> but to take a few steps back, when you're just developing stuff and using the GF shell for quick testing, I'd say it's handier to use spaces
[17:57:48] *** Quits: e3928a3bc2 (~yaaic@2804:14d:5cd8:8447:89d8:e1e2:f7c0:518b) (Ping timeout: 240 seconds)
[17:57:56] <inariksit> so you can define your own function, "myGlue" and at testing phase define it as ++, and when your grammar is finished and ready to be used from another program, then change the definition to glue
[17:58:25] <inariksit> (daherb, what did you say about procrastinating by helping people on #gf? :-D)
[18:01:01] *** Joins: e3928a3bc (~yaaic@2804:18:4805:837:b4d8:dc3a:f327:ad8e)
[18:03:02] <inariksit> e3928a3bc: the latest lines are here, in case you didn't see them https://pastebin.com/VApMmJ07
[18:03:15] <e3928a3bc> thanks for the pointers! I'll take a look at these functions and the article!
[18:04:03] <e3928a3bc> thank you for the pastebin, I'm on the metro now, so connection is terrible
[18:06:33] <inariksit> ^^
[18:08:06] <inariksit> vin-ivar, esg, daherb, how about this for the retreat? :-D https://www.airbnb.com/rooms/19621967?location=Rapla%2C%20Estonia&children=0&infants=0&adults=3&guests=3&check_in=2018-08-18&check_out=2018-08-31&s=99tcu4vS
[18:12:59] <e3928a3bc> as for what you said about the two runtimes, inariksit, is the Haskell one being deprecated?
[18:13:40] <inariksit> the active development happens on the C runtime nowadays
[18:13:55] <inariksit> I don't know if anyone adds anything new to the haskell one
[18:14:07] <inariksit> or fixes bugs
[18:14:51] <e3928a3bc> ok! I'm still far from writing an app or anything, but wanted to have a better view of the environment\n
[18:15:07] <inariksit> but it's not like it will disappear, and aside from the orthography stuff (spaces, capitalisation, ...), the two are pretty much the same
[18:15:21] <inariksit> Haskell one is also better documented
[18:16:07] <inariksit> but anyway I'll use the C runtime for any new project I start, because it feels clearly to have more future
[18:16:11] <inariksit> and it has already more features
[18:16:27] <inariksit> but like, if you had an old app using the haskell, there's no urgency to change it to C
[18:20:23] <e3928a3bc> ok! so for now, no big changes...
[18:20:58] <inariksit> yeah :)
[18:40:34] *** Quits: e3928a3bc (~yaaic@2804:18:4805:837:b4d8:dc3a:f327:ad8e) (Ping timeout: 255 seconds)
[18:56:28] *** Joins: e3928a3bc (~yaaic@2804:18:4804:e85a:ed1c:2571:9217:721e)
[22:23:19] *** Quits: e3928a3bc (~yaaic@2804:18:4804:e85a:ed1c:2571:9217:721e) (Ping timeout: 255 seconds)
[22:49:27] *** Quits: proteus-guy (~proteusgu@180.183.14.107) (Ping timeout: 260 seconds)
[22:55:15] *** Joins: proteus-guy (~proteusgu@180.183.40.34)