[01:31:04] *** Quits: quernd (~user@a95-94-123-120.cpe.netcabo.pt) (Quit: rcirc on GNU Emacs 25.1.1) [03:12:34] *** Quits: JuanDaugherty (~juan@98.4.124.117) (Quit: Ex Chat) [03:41:27] *** Quits: proteus-guy (~proteusgu@14.207.87.4) (Ping timeout: 240 seconds) [03:41:41] *** Joins: proteus-guy (~proteusgu@14.207.87.202) [06:21:24] *** Quits: proteusguy (~proteus-g@2405:9800:bc10:1ca:d126:f9d3:ac1c:b267) (Remote host closed the connection) [06:59:45] *** Joins: proteusguy (~proteus-g@183.88.34.200) [09:17:03] *** Quits: proteusguy (~proteus-g@183.88.34.200) (Ping timeout: 268 seconds) [09:58:18] *** Joins: proteusguy (~proteus-g@2405:9800:bc10:1ca:90f8:7f30:24af:99b1) [10:25:18] *** Joins: venicecat (caa61422@gateway/web/freenode/ip.202.166.20.34) [10:25:34] hello Inari! :o) [10:26:07] For some reason, i keep failing to do a mkV2V [10:26:53] Would it be ok to ask for an example to illustrate what the concrete syntax looks like? [10:27:22] mkV2V : V -> Prep -> Prep -> V2V [10:28:35] sure! [10:28:46] 10qs! [10:29:59] so you have this list of preps http://www.grammaticalframework.org/lib/doc/synopsis.html#Prep [10:30:07] and mkPrep : Str -> Prep in ParadigmsEng [10:30:14] as well as a noPrep [10:30:29] bingo... want = mkV "want" ... i wanted to add a noPrep in there somewhere [10:31:04] okay, so mkV2V (mkV "want") noPrep to_Prep [10:31:20] oh [10:31:31] my brain must have been clogged up [10:31:53] happens easily with GF :-P [10:33:49] :-p [10:34:14] constant not found for no_Prep though [10:34:21] noPrep [10:34:24] not no_Prep [10:36:20] argh cloggy brain! [10:36:37] It works perfectly, thank you for the fish! [10:37:54] ^_^ [11:46:20] *** Quits: venicecat (caa61422@gateway/web/freenode/ip.202.166.20.34) (Ping timeout: 260 seconds) [17:27:02] I'm starting to get pretty happy about this https://github.com/inariksit/GF-testing/blob/master/testLang.sh [17:27:15] esg: you asked some time ago how the testing is going [17:28:54] so now that you change something in your grammar, you can get this list, comparing it to the old grammar: https://github.com/inariksit/GF-testing/blob/master/data/dutch-category-changes.md [17:29:44] then you can generate examples using a given function (DetNP in the following example), and also the program automatically generates contexts for those examples [17:29:47] like this: https://github.com/inariksit/GF-testing/blob/master/data/dutch-prep-np-contraction.md#new-grammar-1 [17:32:51] the ideal workflow is described here: https://github.com/inariksit/GF-testing/blob/master/testLang.sh#L19-L36 [17:33:27] also you can test just single words, if you changed a constructor or something: https://gist.github.com/inariksit/b892acfc46b1d30da548124ae48f3bba [18:33:12] *** Joins: JuanDaugherty (~juan@98.4.124.117) [19:09:31] *** Joins: e3928a3bc (~user@2804:14d:5cd8:8447::3) [19:09:35] hey people [19:09:49] I'm having trouble compiling GF with the C runtime support [19:10:09] I'm following the tutorial exactly http://www.grammaticalframework.org/doc/gf-developers.html#toc16 [19:12:15] the error I'm getting https://pastebin.com/raw/WNsHJCWQ [20:09:55] so you have installed the C runtime first? [20:10:11] then you just need to find where it put the files, and give those directories as arguments to --extra-include-dirs= and --extra-lib-dirs= [20:11:19] for me it's /usr/local/include and /usr/local/lib [20:24:27] oh, that might be it [20:24:41] I'll try it [20:26:16] inariksit: I also have another weird issue, could take a look at it please? I've written it up here https://pastebin.com/raw/N8yXvqvB [20:27:13] e3928a3bc: I'll have a look at it later, I'm busy now! :-) [20:27:30] ok, thanks a lot, inariksit!! [21:36:46] adding the flags helped it continue, but it found more errors later on [21:36:56] in the haskell code [21:37:17] I'm updating my repo to see if that fixes it [21:38:12] is the master branch stable? [21:40:49] I imagine so [21:41:14] like, I would imagine that krasimir doesn't push stuff that is totally half-baked [21:41:28] and yeah there is actually a branch called c-runtime [21:49:10] okay now I'm free, I looked at your file [21:49:28] can you give me your grammar and I can test it on my setup? [21:51:17] about the Ints, I've never used GF for other than natural language, so I have no experience on that kind of behaviour [21:56:47] I'd think so too! [21:57:01] but it doesn't seem to work :/ I must be doing something wrong.. [21:57:15] the grammar's here: https://github.com/odanoburu/gf-data [21:57:25] inariksit: thank you again! [22:00:08] I picked the ints as an example, but strings are also causing problems... (String, not Str) [22:00:24] but none are used in traditional grammars, right? [22:01:00] I think String might be used for parsing words outside lexicon [22:04:46] oh btw, now that I'm trying to test your grammar, I'm getting this error myself :-D [22:04:51] Ambiguous interface for ‘PGF2’: [22:04:52] it was found in multiple packages: pgf2-0.1.0.0 pgf2-bind-0.1.0.0 [22:05:15] have you run into it? [22:11:54] nope [22:12:13] I think I'd have PGF2 installed only if I had manage to compile the C runtime :P [22:12:31] (I'm using the .deb file shell with C support, but no PGF2 module) [22:13:49] okay now it works, I'm getting the same behaviour: [true,false,null] parses but [1,2,3] not [22:14:31] funny, isn't it? [22:14:57] if you use cc in the haskell runtime it seems to be ok [22:15:03] it should parse (i think) [22:24:57] yeah, it looks like it should [22:26:08] so why do the atomStr and mkInt functions need to be from Ints and Strings? [22:27:41] so that I can parse arbitrary strings and numbers! is there a better way? [22:28:04] yeah, that's a good reason :-D hmm I wonder if the Symbol type could be useful [22:28:37] as I said, I have no experience in parsing arbitrary stuff in GF, so take it with a grain of salt :-P [22:28:44] the one from the Symbolic module? [22:29:17] should I send these problems to the mailing list? (the installation and this one?) [22:29:26] yeah, do it! [22:29:42] did you search for the installation problem? [22:29:46] I've seen it tons of times before [22:30:03] but it might be just on individual people's computers and never on the mailing list [22:31:38] not really, there are too many errors T.T [22:32:16] I'm writing a Dockerfile so that no one will have these problems anymore -- except I can't get it to work in the first place :P [22:32:33] https://github.com/odanoburu/GF/blob/docs/Dockerfile [22:32:39] there you can see the steps I'm taking [22:32:54] the base system is Debian 9 with ghc 8.2.1 [22:32:54] nice ^^ [22:33:36] oh yeah right you said you had other errors too [22:33:43] so yeah definitely write to the list! [22:35:27] the error log: https://pastebin.com/raw/1kvBzu5i [22:43:27] alright, I haven't seen that before [22:51:40] using an older version of ghc gives less errors, but still https://pastebin.com/raw/zCrQLRPq [22:52:06] mind you that I'm using the code from the commit of the GF 3.9 release [22:52:12] so I must be doing something wrong [22:52:56] I don't know, it seems that dist/build/gf/gf: error while loading shared libraries: libpgf.so.0: cannot open shared object file: No such file or directory is the key thing [22:53:03] I've never tried the -cshell option [22:53:26] I just install GF in the old way, and then I separately install the c-runtime, and after that the haskell bindings [22:53:54] and then I parse anything I want with BIND or SOFT_BIND using an external program [22:54:05] but yeah even that didn't work with your grammar [22:57:26] that makes sense, although the haskell errors (or is it warnings?) probably won't go away [22:57:58] they are warnings up to that point [22:58:03] how do you do it? (the old way you refer is which one?) [22:58:14] just cabal install [22:58:21] for GF I mean [22:58:22] :-D [22:58:28] which external program?? [22:58:42] oh right [22:58:43] and then in the GF/src/runtime/c/ there's instructions in INSTALL or osmething [22:59:00] I thought you had to do it in the order it says there [22:59:07] (in the tutorial) [22:59:23] I'll try it! [22:59:33] which ghc do you use? [23:00:30] as for the external program, just something like this: https://gist.github.com/inariksit/4591fe1dabfad3fbf348846ce7c899ea [23:00:43] I have 8.0.2 [23:01:10] oh right, so using the haskell bindings.. [23:01:26] well that was the thing I had the new error message [23:01:41] and I didn't [23:01:42] I see [23:02:25] okay, apparently if I change the first import to import "pgf2-bind" PGF2, then ParseOk doesn't work [23:02:27] I'd rather have the c runtime work too, though, because that's more portable if I don't need haskell to do any pre/post processing [23:02:42] but this method you provided should also install it, right? [23:03:00] yeah certainly, it's the first step to getting any bindings [23:03:20] in src/runtime you have bunch of folders for bindings in different programming languages [23:05:37] great! [23:05:42] I'm trying it! [23:11:23] *** Quits: JuanDaugherty (~juan@98.4.124.117) (Quit: Ex Chat) [23:53:54] inariksit: no deal :/ [23:54:03] it says there's a missing C library [23:54:12] although it seems to have been built [23:54:32] and I'm passing `cabal install` the path to it [23:56:02] libgu and libpgf, or some other C library? [23:57:49] sg! [23:57:57] it is there