[00:04:12] <wmacmil> sorry, i was so tired last night i forgot to paste it .
[00:04:14] <wmacmil>   ModifyCommand conj cond com =  { ut = <(C.mkAdv conj <(mkUtt cond) : Adv> <({s = com.ut} ) : Adv>) : Utt>} ** { adv = ""} ;
[00:04:21] <wmacmil> the most recent incarnation
[00:07:43] <wmacmil> but i don't understand how the coercions work, the following gives a lock error
[00:07:44] <wmacmil> oper
[00:07:44] <wmacmil>   advIsUtt : Adv -> Utt ;
[00:07:44] <wmacmil>   advIsUtt a = <a : Utt> ;
[00:07:54] <wmacmil>       Warning: missing lock field lock_Utt
[00:08:17] <wmacmil> and i didnt see them documented in the book
[00:10:02] <daherb> did tu try "lin Utt" in front of it?
[01:31:23] <wmacmil> @daherb, in front of what? 
[03:50:09] <inariksit> advIsUtt a = lin Utt a
[03:51:02] <inariksit> yeah, it's a bit confusing with the different syntaxes
[03:51:26] <inariksit> I use <foo : Bar> when I'm giving foo as an argument to something that expects Bar
[03:51:55] <inariksit> but in your case, you have an oper that just coerces Foo to Bar
[03:52:18] <inariksit> then we need to actually add the lock_Bar field to the Foo
[03:52:29] <inariksit> so that's when to use lin Bar foo
[03:52:42] <inariksit> lin Bar {s = "hello"}
[03:52:46] <inariksit> is exactly the same as
[03:53:12] <inariksit> {s = "hello" ; lock_Bar = <>}
[04:18:09] <inariksit> if you're using them regularly, and not just one-off in a single function, then I'd make helper opers
[04:18:20] <inariksit> example following
[04:50:12] <inariksit> https://gist.github.com/inariksit/0f9902873d18068a477e96ddd71f6561
[04:50:24] <inariksit> got distracted with my actual work :-D
[04:52:34] <inariksit> well actually, providing GF documentation is also a real part of my job!
[04:52:54] <inariksit> I expect to copy and paste the latest logs from here into a blog post at some point
[05:44:21] <wmacmil> tack, inari
[05:44:31] <wmacmil> tack som alltid
[05:45:13] <inariksit> hope it helps! the previous hints were not really the complete picture
[05:45:35] <inariksit> I think I usually use the <foo : Bar> hack when  just want to return a Bar
[05:45:46] <inariksit> and not actually in the end return another Foo
[05:46:02] <inariksit> so I didn't think of explaining the lin Foo thing
[06:50:00] <wmacmil> now i'm getting a type error
[06:50:08] <wmacmil> when i try to adjust your code to mine
[06:50:09] <wmacmil>        type of and_Conj
[06:50:09] <wmacmil>       expected: Str
[06:50:09] <wmacmil>       inferred: {lock_Conj : {}; n : ParamX.Number; s1 : Str; s2 : Str}
[06:51:06] <wmacmil>   ModifyCommand conj cl imp =
[06:51:07] <wmacmil>       let clUtt : Utt = mkUtt cl ;
[06:51:07] <wmacmil>           impUtt : Utt = mkUtt imp ;
[06:51:07] <wmacmil>           fakeAdv : Adv =
[06:51:07] <wmacmil>             mkAdv conj 
[06:51:08] <wmacmil>                   <clUtt : Adv> 
[06:51:10] <wmacmil>                   <impUtt : Adv> ;
[06:51:12] <wmacmil>       in lin Utt fakeAdv ; 
[06:51:41] <wmacmil> sorry with the and_Conj replaced
[06:56:32] <wmacmil> nvm i think it was calling the wrong mkAdv
[07:00:35] <wmacmil> it finally works, tack
[07:02:24] <inariksit> yay!
[09:00:14] *** Quits: wmacmil (~wmacmil@2600:1700:cd90:8f10:340c:b145:c6ee:fdbd) (Quit: Leaving)
[22:24:16] *** Quits: step21 (~step21@apus.uberspace.de) (Read error: Connection reset by peer)
[22:27:13] *** Joins: step21 (~step21@apus.uberspace.de)