[06:52:19] *** Quits: proteusguy (~proteus-g@2405:9800:bc10:1ca:9db1:e95b:29b4:c405) (Remote host closed the connection) [08:38:02] *** Joins: proteusguy (~proteus-g@180.183.119.155) [12:57:25] *** Quits: proteusguy (~proteus-g@180.183.119.155) (Remote host closed the connection) [15:05:37] *** Joins: e3928a3bc (~user@189.125.124.24) [15:10:16] hello people! [15:10:17] [15:10:17] I'm having trouble embedding grammars in Haskell. on this computer I have installed GF using the .deb binary, but I don't seem to be able to `import PGF` on my haskell programs. is this a $PATH problem, or do I have to install the haskell bindings separately? [15:19:38] *** Joins: proteusguy (~proteus-g@2405:9800:bc10:1ca:f0c2:ca32:6b5d:c003) [15:29:54] e3928a3bc: the deb file should also contain the haskell binding (but maybe not for your haskell version? it seems to be installed in /usr/lib/x86_64-linux-ghc-8.0.2/) [15:32:13] daherb: you are right, thank you! I was looking at /usr/share/x86_64-linux-ghc-8.0.2/gf-3.9/, not at /usr/lib... [15:35:37] what's the difference between PGF and PGF2? [15:40:33] PGF2 is the haskell binding to the new C runtime [15:42:26] *** Joins: JuanDaugherty (~juan@98.4.124.117) [15:54:44] oh, right. so I should probably be using it, right? [15:56:46] for most practical purposes both should work [15:56:54] but some stuff only works in PGF2 [15:57:15] like if you're using a grammar with a lot of BIND tokens, or fancy capitalisation rules, or SOFT_SPACE [15:57:18] then use the C runtime [15:57:27] but if you're doing simple things with foods grammar, any of them will do [15:58:04] inariksit: ok! thank you for the info! [15:58:16] but, uhm, it doesn't work out of the box. I included that path in my $PATH, but maybe I still have to include the path in some cabal config file? [15:59:03] e3928a3bc: and some things only work with the bindings to the haskell runtime, so use that one if you want to do e.g. random generation. [15:59:25] "that one" = PGF [16:00:08] e3928a3bc: I don't know, I use a mac and everything works for me [16:00:50] I don't think you should add anything to some cabal config files [16:01:09] it's just a normal haskell library [16:01:13] the PGF [16:01:16] (and PGF2) [16:02:40] esg: great! thanks! [16:03:24] inariksit: I thought it might have been the fact that I installed gf before haskell on this computer, but I've reinstalled and no deal [16:03:37] okay [16:04:05] do you have the GF source code? you could try to install PGF2 and see if that works [16:04:29] go to src/runtime/haskell-bind/ and cabal install [16:04:53] if I remember correctly, PGF comes by default but PGF2 you have to install yourself [16:05:02] so just see if that makes any difference [16:08:17] not on this computer, no! I'll try on my other one later! [16:08:17] [16:08:17] that's related to one of those questions I asked on a friday night, haha the GF repo is too big for both my internet and processor speeds. running git status takes ages. [16:08:17] [16:08:20] I was going to propose on the gf-dev list using git submodules like they do for GHC https://ghc.haskell.org/trac/ghc/wiki/WorkingConventions/Git/Submodules, and splitting the repo! [16:08:23] [16:08:26] I could help, too! [16:09:07] brb [16:11:32] mm I don't think that will happen anytime soon :/ like there's this pull request by mbrock https://github.com/GrammaticalFramework/GF/pull/14 [16:12:03] *** Joins: e3928a3bc2 (~yaaic@201.39.147.24) [16:12:17] he did the whole thing already, but krasimir just commented a long time ago that he'd have to confirm himself that this is a legit thing that works and stuff, and he doesn't have time for those things [16:12:35] which is understandable, because he's instead developing PGF2 and other stuff more directly related to GF [16:13:12] so I'd be surprised if the core GF developers would either themselves split the repo into submodules, or let someone else do it [16:13:33] but in any case, do write on the list! these things may take time [16:13:41] but maybe in 2020 we'll have travis and submodules and whatnot :-D [16:16:08] *** Quits: e3928a3bc2 (~yaaic@201.39.147.24) (Ping timeout: 240 seconds) [16:16:41] :-) [16:16:47] *** Joins: e3928a3bc2 (~yaaic@201.39.147.24) [16:18:52] *** Joins: e3928a3bc3 (~yaaic@201.39.147.24) [16:18:52] *** Quits: e3928a3bc2 (~yaaic@201.39.147.24) (Read error: Connection reset by peer) [16:19:00] geez [16:19:09] sorry, connectivity problems! [16:27:45] *** Quits: e3928a3bc3 (~yaaic@201.39.147.24) (Ping timeout: 248 seconds) [17:12:01] inariksit: that's sad :/ I mean, we'd all have to learn a few new git commands, which is bad -- but I'd say it probably works, as people already use it.. [17:13:02] we can even keep the commit history while doing it! and there's no need to touch the old repo until everyone is happy with the result [17:13:21] but I get it's a big change and there's more important stuff to do! [17:14:06] (I might have missed some messages, if so do tell me!) [17:21:14] no you didn't [17:21:43] I think probably the most feasible thing would be to get someone in the core team, preferably here in gothenburg, who really has time and would already know these things [17:22:29] I mean a new person, instead of waiting for krasimir and thomas and aarne to catch up with git [17:22:48] if digital grammars keeps hiring, maybe they could even pay for someone to be responsible for just that [17:22:53] but this is 100% speculation [17:23:15] and maybe those things are not even that difficult, and aarne, krasimir and thomas would learn them in 10 minutes, but they just think it will take a year [17:25:40] I'm sure they would! I guess I'll wait, then.. [17:31:10] I mean, do send the email to the list! maybe I'm wrong and they all have time and energy now that it's christmas holiday soon [17:31:23] and even if they don't have the time now, it's still good for them to see that there's interest in the community [17:31:29] to help us improve [18:19:45] ok! I'll try! [18:20:38] I installed gf using haskell now, and now I can import PGF! but not PGF2, for some reason haha [18:20:49] but that's enough for me, for the time being [18:44:14] hum, that's embarassing. I can import PGF, but when I compile the file, I got: PGF.hs:156:1: error: [18:44:14] Failed to load interface for ‘PGF.CId’ [18:44:14] it is a hidden module in the package ‘gf-3.9’ [18:44:14] Use -v to see a list of the files searched for. [19:08:41] *** Quits: e3928a3bc (~user@189.125.124.24) (Read error: Connection reset by peer)