[00:36:12] *** Quits: Gurkenglas (~Gurkengla@dslb-188-103-074-033.188.103.pools.vodafone-ip.de) (Ping timeout: 272 seconds)
[01:20:18] *** Joins: Gurkenglas (~Gurkengla@dslb-188-103-074-033.188.103.pools.vodafone-ip.de)
[02:51:01] <Gurkenglas> &querz begiak
[02:51:29] <Gurkenglas> @help
[02:53:32] <Gurkenglas> Ah well, it reached him last time. inariksit, can I gm make all the files in a folder (specifically this one https://github.com/GrammaticalFramework/GF/tree/master/lib/src/english ) without enumerating them myself?
[02:53:33] <begiak> [ GF/lib/src/english at master · GrammaticalFramework/GF · GitHub ]
[02:54:47] <Gurkenglas> (Or is that a bad idea for some reason?)
[03:13:01] *** Quits: Gurkenglas (~Gurkengla@dslb-188-103-074-033.188.103.pools.vodafone-ip.de) (Ping timeout: 244 seconds)
[03:33:22] *** Joins: Gurkenglas (Gurkenglas@dslb-188-103-074-033.188.103.pools.vodafone-ip.de)
[04:39:27] *** Joins: daherb (~daherb@ppp-93-104-178-242.dynamic.mnet-online.de)
[04:42:34] *** Quits: daherb_ (~daherb@ppp-93-104-178-26.dynamic.mnet-online.de) (Ping timeout: 255 seconds)
[05:04:18] *** Quits: Gurkenglas (Gurkenglas@dslb-188-103-074-033.188.103.pools.vodafone-ip.de) (Ping timeout: 255 seconds)
[08:23:27] <begiak> apertium: unhammer * 61101: /trunk/apertium-nno-nob/configure.ac: require apertium-lex-tools
[08:23:35] <begiak> apertium: unhammer * 61102: /trunk/apertium-sme-nob/apertium-sme-nob.sme-nob.t1x: todo→done
[08:23:44] <begiak> apertium: unhammer * 61103: /trunk/apertium-sme-nob/configure.ac: don't require xsltproc; use new compact ac_path thing
[10:24:08] *** Joins: spectre (~fran@c-68-54-130-222.hsd1.in.comcast.net)
[12:15:11] <inariksit> begiak: tell Gurkenglas that you can do in general,   gf *.gf  and it will compile all .gf files; but you don't really need to do that, LangEng.gf already consists of all the other modules
[12:15:11] <begiak> inariksit: I'll pass that on when Gurkenglas is around.
[12:16:00] <inariksit> begiak: tell Gurkenglas but if you want to have several LangXxx, then you just need to do gf -make english/LangEng.gf german/LangGer.gf
[12:16:00] <begiak> inariksit: I'll pass that on when Gurkenglas is around.
[12:21:40] *** Joins: Gurkenglas (Gurkenglas@dslb-188-103-074-033.188.103.pools.vodafone-ip.de)
[12:21:40] <begiak> Gurkenglas: You have messages.
[12:21:54] <inariksit> ooh hey, for once we're online at the same time! :-D
[12:48:50] <begiak> apertium: jezral * 61104: /branches/packaging/trunk/apertium-nno-nob/debian/control: nno-nob depends
[13:17:59] <vin-ivar> inariksit,
[13:18:26] <vin-ivar> how/where do we get the keys to the flat?
[13:19:14] <inariksit> vin-ivar: I should contact the guy again
[13:19:21] <inariksit> when I booked it, nobody had flights yet
[13:19:28] <inariksit> so I just said I'll contact him later
[13:19:56] <vin-ivar> ah okay, cool
[13:20:02] <vin-ivar> they might ask me that in my interview
[13:25:25] <Gurkenglas> @messages
[13:25:25] <begiak> Gurkenglas: 10:15Z <inariksit> tell Gurkenglas that you can do in general,   gf *.gf  and it will compile all .gf files; but you don't really need to do that, LangEng.gf already consists of all the other modules
[13:25:27] <begiak> Gurkenglas: 10:16Z <inariksit> tell Gurkenglas but if you want to have several LangXxx, then you just need to do gf -make english/LangEng.gf german/LangGer.gf
[13:27:03] <Gurkenglas> Kay shall try that. Nah we aren't, this PC is just also used by another. Aaaand I'm gone again.
[14:03:39] <inariksit> hahah okay
[14:04:18] <inariksit> Gurkenglas: also if you want to leave messages with begiak, just do this
[14:04:25] <inariksit> begiak: tell Gurkenglas this is a test
[14:04:25] <begiak> inariksit: I'll pass that on when Gurkenglas is around.
[14:04:33] <inariksit> begiak: tell inariksit this is another test
[14:04:33] <begiak> You can tell yourself that.
[14:04:36] <inariksit> :-D
[14:04:54] <inariksit> begiak: do you like particles
[14:04:55] <begiak> inariksit: this is my particles face :(((((
[15:06:37] <begiak> apertium: vitaka * 61105: /branches/apertium-transfer-tools-generalisation-hbs/phrase-extraction/transfer-tools-scripts/apertium-hbs_SR-hbs_HR.transfer-at.hbs_SR-hbs_HR.atx (+): added missing files
[15:14:03] <begiak> apertium: vitaka * 61106: /branches/apertium-transfer-tools-generalisation-hbs/extract-rules.sh: added missing files
[15:14:12] <begiak> apertium: vitaka * 61107: /branches/apertium-transfer-tools-generalisation-hbs/extract-rules.sh: added missing files
[15:26:35] <begiak> apertium: vitaka * 61108: /branches/apertium-transfer-tools-generalisation-hbs/phrase-extraction/transfer-tools-scripts/Makefile: chapuza tt1
[15:31:36] <begiak> apertium: vitaka * 61109: /branches/apertium-transfer-tools-generalisation-hbs/phrase-extraction/transfer-tools-scripts/Makefile: chapuza tt1
[15:36:43] <begiak> apertium: vitaka * 61110: /branches/apertium-transfer-tools-generalisation-hbs/phrase-extraction/transfer-tools-scripts/Makefile: chapuza tt1
[15:44:07] <begiak> apertium: vitaka * 61111: /branches/apertium-transfer-tools-generalisation-hbs/phrase-extraction/transfer-tools-scripts/Makefile: chapuza tt1
[16:24:13] <begiak> apertium: vitaka * 61112: /branches/apertium-transfer-tools-generalisation-hbs/phrase-extraction/transfer-tools-scripts/apertium-hbs_SR-hbs_HR.transfer-at.hbs_SR-hbs_HR.atx: cambia lex tags
[16:27:43] *** Quits: Gurkenglas (Gurkenglas@dslb-188-103-074-033.188.103.pools.vodafone-ip.de) (Ping timeout: 255 seconds)
[16:31:43] <begiak> vislcg3: tino * 10815: /tools/vislcg3/trunk/: src/version.hpp, ChangeLog, src/GrammarApplicator_runRules.cpp: Fix removing enclosure owner
[16:43:58] *** Joins: Gurkenglas (Gurkenglas@dslb-188-103-074-033.188.103.pools.vodafone-ip.de)
[16:43:59] <begiak> Gurkenglas: You have messages.
[16:47:48] <Gurkenglas> @messages
[16:47:48] <begiak> Gurkenglas: 12:04Z <inariksit> tell Gurkenglas this is a test
[17:38:24] <Gurkenglas> inariksit, http://pastebin.com/VxzXciQq doesn't sound like it should happen
[17:38:25] <begiak> [ C:\Users\Gurkenglas\Desktop\Haskell\gf-3.6>gf -make src/english/LangEng.gf - co - Pastebin.com ]
[17:45:28] <inariksit> Gurkenglas: try to download GF 3.7
[17:45:33] <inariksit> maybe it's fixed there
[17:45:39] <inariksit> it was just released
[17:46:29] <inariksit> did you download the GF binary or have you also compiled it yourself on your windows?
[17:46:50] <inariksit> well that doesn't really make a difference, this is just a RGL problem :-P
[17:46:53] <Gurkenglas> I downloaded the binary, and copied the src from the zip I downloaded from the github page
[17:47:00] <inariksit> okay
[17:47:44] <Gurkenglas> I copied part of the lib from that zip, then aborted when I noticed there was already a folder of that name in there, maybe it screwed something up. But it sounds likely that the github zip is just using features from 3.7
[17:48:11] <inariksit> yeah, it's likely that the Predef that comes in the binary doesn't have CAPIT
[17:48:29] <inariksit> you can go and compile the 3.7 files and then move them manually to your GF_LIB_PATH
[17:48:38] <inariksit> or just download the 3.7 binary
[17:49:20] <inariksit> ok the latter is obviously better solution :-D I'm just used to suggesting things to people who don't want to compile stuff themselves and there are new commits in the RGL that are not compatible with the binaries
[17:49:23] <Gurkenglas> I downloaded http://www.grammaticalframework.org/download/gf-3.7-bin-windows.zip . I am planning to now copy the src folder from the github zip into the 3.7 bin and try gf make
[17:49:25] <Gurkenglas> *-
[17:49:33] <inariksit> sounds good
[17:50:29] <inariksit> if you still have problems after that, check that you have $GF_LIB_PATH specified correctly, ie. that it  points to the 3.7 files
[17:50:32] <inariksit> http://www.grammaticalframework.org/~inari/gf-windows.html#toc3
[17:55:04] <Gurkenglas> Got a Lang.pgf, gonna try it with Translate.hs.  Though it will be uninteresting, since I only used LangEng. (There were also warnings. http://pastebin.com/1YrZdNVX ) src/german/LangGer.gf did not work: http://pastebin.com/Eq9N9Vz3
[17:55:05] <begiak> [ C:\Users\Gurkenglas\Desktop\Haskell\gf-3.7\bin>gf -make src/english/LangEng.gf s - Pastebin.com ]
[18:00:15] <inariksit> the warnings are ok
[18:00:25] <Gurkenglas> (Deleting the 3.6 folder had no effect, so library mismatches are improbable.)
[18:00:39] <inariksit> the Construction module is not very important anyway, it's just some experimental stuff :-P
[18:00:52] <inariksit> as for the missing Prelude.gf, set up a GF_LIB_PATH
[18:00:56] <inariksit> http://www.grammaticalframework.org/~inari/gf-windows.html#toc3
[18:01:55] <Gurkenglas> Wrongly dismissed your 17:50 thing then I guess :I *tries*
[18:02:42] <inariksit> it's strange though that you are able to compile LangEng from gf-3.7/bin/
[18:02:44] <inariksit> but not LangGer
[18:03:11] <inariksit> can you just make src/german/LangGer.gf on its own?
[18:03:23] <inariksit> then if you have two PGF files, you can put them to a same place and merge them into one
[18:03:23] <Gurkenglas> I tried, and the result is included in the pastebin
[18:03:29] <inariksit> aah right
[18:03:31] <inariksit> yeah
[18:03:35] <inariksit> that's strange
[18:03:57] <inariksit> does it help if you actually go to the directory and then try gf -make?
[18:05:27] <inariksit> ah
[18:05:45] <inariksit> add to the LangGer.gf file this:
[18:05:52] <inariksit> --# -path=.:../abstract:../common:../api:../prelude
[18:05:55] <Gurkenglas> Without having added it to the path: http://pastebin.com/k7LDXnGA
[18:05:56] <begiak> [ C:\Users\Gurkenglas\Desktop\Haskell\gf-3.7\bin\src\german>C:\Users\Gurkenglas\De - Pastebin.com ]
[18:06:40] <inariksit> yeah, now I got what was wrong, so false alarm :) well it's going to be useful anyway to have the GF_LIB_PATH
[18:07:02] <inariksit> but the real reason must be that there is no --# -path=../prelude  in LangGer.gf
[18:07:11] <Gurkenglas> Worked.
[18:07:13] <inariksit> \o/
[18:07:20] <inariksit> I will push that
[18:07:55] <Gurkenglas> Should there be automatic testing for making all the standard Lang*.gfs?
[18:08:39] <inariksit> if you type make in the lib/src directory, it does that
[18:09:26] <inariksit> ah sorry now I got what you meant
[18:09:38] <inariksit> I'm not sure why does it work without that too
[18:09:56] <inariksit> maybe the developers just have some magic environment variables :-P
[18:10:08] <inariksit> or gf_lib_path does that too
[18:14:39] <Gurkenglas> "Translate: src/runtime/haskell/PGF/Binary.hs:(28,12)-(30,46): Non-exhaustive patterns in case
[18:14:39] <Gurkenglas> ". Looking at https://github.com/GrammaticalFramework/GF/blob/master/src/runtime/haskell/PGF/Binary.hs and scratching my head at "(28,12)-(30,46)", which interpreted as line/column coordinated dont seem to point to anything.
[18:14:40] <begiak> [ GF/Binary.hs at master · GrammaticalFramework/GF · GitHub ]
[18:16:44] <inariksit> shit shit
[18:16:45] <inariksit> hmm
[18:17:02] <inariksit> does that come from when you're trying to open the pgf file in a haskell program?
[18:17:32] <Gurkenglas> Yep, this one: http://lpaste.net/135390
[18:17:33] <begiak> [ No title :: lpaste — Lambda pastebin ]
[18:18:18] <inariksit> :/
[18:19:11] <inariksit> you can also try the C-based PGF thing
[18:19:34] <inariksit> it's in src/runtime/haskell-bind
[18:19:36] <Gurkenglas> ( The one from http://www.grammaticalframework.org/doc/tutorial/gf-tutorial.html#toc148 )
[18:19:59] <inariksit> ok that code is most likely super old
[18:21:44] <inariksit> if you try the code in haskell-bind (actually the C runtime is the one that is more actively developed nowadays) and these examples https://github.com/GrammaticalFramework/GF/tree/master/src/runtime/haskell-bind/examples
[18:21:45] <begiak> [ GF/src/runtime/haskell-bind/examples at master · GrammaticalFramework/GF · GitHub ]
[18:21:48] <inariksit> do they work?
[18:22:03] <inariksit> the api for the haskell bindings is almost the same as for the original haskell runtime
[18:22:11] <inariksit> the code is just not well documented
[18:27:40] <Gurkenglas> (To "super old": http://lpaste.net/135391 (assembled from reading hackage) yields "*** Exception: src/runtime/haskell/PGF/Binary.hs:(28,12)-(30,46): Non-exhaustive patterns in case
[18:27:40] <Gurkenglas> ".)
[18:27:40] <begiak> [ No title :: lpaste — Lambda pastebin ]
[18:30:04] <inariksit> can you do even any simpler operations to the PGF file? like languages
[18:30:25] <inariksit> or startCat
[18:30:36] <inariksit> any of these https://hackage.haskell.org/package/gf-3.6/docs/PGF.html#g:3
[18:31:33] <inariksit> also, you're trying to parse something in the category "This guy is a phony!"
[18:31:45] <Gurkenglas> main = readPGF "GerEng.pgf" >>= print . languages produces the same
[18:31:50] <inariksit> okay
[18:32:28] <Gurkenglas> (putting readPGF "GerEng.pgf" into ghci produces the same)
[18:32:31] <inariksit> I can't really help in guessing what might be wrong in the PGF library, I know krasimir doesn't actively develop it anymore
[18:32:41] <inariksit> so I would suggest try the PGF2
[18:32:46] <inariksit> https://github.com/GrammaticalFramework/GF/tree/master/src/runtime/haskell-bind
[18:32:47] <begiak> [ GF/src/runtime/haskell-bind at master · GrammaticalFramework/GF · GitHub ]
[18:34:43] <Gurkenglas> "HOW TO COMPILE: cabal install" <- cabal install what? cabal install PGF2 isn't known. Surely the package isn't called haskell-bind?
[18:35:32] <inariksit> pgf2-bind seems to be the name
[18:35:37] <inariksit> in the .cabal file
[18:36:27] <Gurkenglas> http://pastebin.com/P6qbg9WY
[18:36:27] <begiak> [ gurkenglas@gurkenglas-VirtualBox:/media/sf_Haskell/DialogeMitComputern$ cabal up - Pastebin.com ]
[18:36:50] <inariksit> yeah, just go to the directory and type cabal install
[18:37:21] <inariksit> if that doesn't work, then I suggest you just send email to the GF mailing list
[18:37:54] <inariksit> I can also send you a PGF file that works on my computer, if you want to make sure it's not about the file itself
[18:38:12] <inariksit> also the fact that you're using windows is likely to be a cause of problems
[18:38:33] <inariksit> nobody of the developers is using windows, so there might be some surprising things
[18:38:45] <inariksit> it's not this difficult usually! :-P
[18:39:04] <Gurkenglas> Nope, I couldn't even cabal install gf on Windows, it wouldn't install network due to some cygwin/mingw confusion I was told
[18:39:10] <inariksit> >___>
[18:39:11] <inariksit> okay
[18:39:14] <inariksit> shit that sucks
[18:39:21] <Gurkenglas> I'm doing all the gf stuff on an Ubuntu 14.04 vm I had lying around
[18:39:25] <inariksit> aah okay
[18:39:46] <inariksit> but you had paths like C:\something in the pastes?
[18:40:08] <vin-ivar> WINDOWS
[18:40:20] <vin-ivar> it must be a pain to set up all of this windows
[18:40:34] <Gurkenglas> I mean, the haskell gf stuff. The gf make stuff I did on windows ._.
[18:41:00] <Gurkenglas> Someone have mercy and put a warning on the gf page for anyone who tries to download the windows installer
[18:41:59] <inariksit> okay
[18:42:02] <inariksit> hehe yeah
[18:42:18] <inariksit> nobody really wants to support windows
[18:42:36] <inariksit> but can you just do everything on your virtual machine?
[18:42:48] <inariksit> http://www.grammaticalframework.org/~inari/LangSpa.pgf also here's another pgf to try
[18:43:42] <inariksit> I'm sorry I can't help more :/ https://groups.google.com/forum/#!forum/gf-dev is my next suggestion
[18:44:04] <Gurkenglas> Also doesn't work, which means that moving everything to the vm shouldn't solve the problem
[18:44:43] <Gurkenglas> Oh wait the vm still has gf 3.6.
[18:45:43] <inariksit> it might well be also that 3.7 was buggy to start with
[18:45:57] <inariksit> in which case it's valuable information if you post your error messages to the list
[18:46:03] <inariksit> the developers are not here
[18:46:06] <inariksit> on the channel
[18:46:35] <Gurkenglas> If I get the vm to compile a working program from a .hs, can I also generate something that any computer, even windows, will be able to turn into a working executable?
[18:46:59] <Gurkenglas> How about linking them to this irc conversation?
[18:47:45] <inariksit> this is going to be a loong conversation :-D
[18:47:52] <inariksit> just to summarise the main points rather
[18:48:58] <inariksit> but anyway, PGFs should work the same regardless of the OS
[18:49:42] <Gurkenglas> I mean, can I output something that doesn't require the user to have gf or even haskell installed?
[18:50:28] <vin-ivar> you can compile .hs files to executables, iirc
[18:50:41] <inariksit> yeah; for instance there are android apps that use GF
[18:51:28] <inariksit> but I've never tried that, don't know how that works
[18:51:55] <Gurkenglas> Why does cabal install gf produce so many warnings? These deprecation stuff sounds really easy to fix
[18:52:05] <Gurkenglas> *This
[18:52:29] <inariksit> hehe, I don't know :-P thomas hallgren and krasimir angelov are the main developers
[18:53:11] <vin-ivar> Gurkenglas, changes in the dependencies mainly
[18:53:18] <inariksit> just out of curiosity, could you try to install the PGF2 also on your vm?
[18:53:23] <vin-ivar> they're probably easy to fix but annoying and not super necessary (not sure though)
[18:53:52] <Gurkenglas> Was gonna do that, I think that got lost at some point. Why do you say "out of curiosity"?
[18:58:50] <Gurkenglas> Installed new gf version, *** Exception: Unsupported PGF version (31522,28257). Failed reading at byte position 4 from trying to readPGF your LangSpa.pgf and http://pastebin.com/AafBKHZy from the GerEng.pgf. Progress, I guess?
[18:58:55] <begiak> [ gurkenglas@gurkenglas-VirtualBox:/media/sf_Haskell/DialogeMitComputern$ ghci Tra - Pastebin.com ]
[18:59:36] <inariksit> Gurkenglas: yeah, that was compiled with old GF
[18:59:46] <inariksit> so not working is expected ^_^
[18:59:57] <Gurkenglas> ("Ich bin ein " is a prefix of many a sentence.)
[19:00:11] <inariksit> now compile a new PGF and try to read that!
[19:00:27] <Gurkenglas> I did, GerEng. See the paste.
[19:01:31] <inariksit> aah ok, cool
[19:02:37] <inariksit> try with lowercase ich?
[19:03:22] <inariksit> is that Translate.hs the same as in http://www.grammaticalframework.org/doc/tutorial/gf-tutorial.html#toc148 ?
[19:04:01] <Gurkenglas> yes
[19:04:05] <inariksit> okay
[19:04:40] <inariksit> yeah okay so that just seems that it exits if there is no parse
[19:04:51] <inariksit> so GF is quite brittle, e.g. with case sensitivity
[19:05:20] <inariksit> intended behaviour is that the end users should just write their own pre- and post-processing
[19:08:32] <Gurkenglas> Can you recommend a sentence that works? "i exist" and "I exist" don't.
[19:08:47] <inariksit> I am an apple
[19:09:24] <inariksit> Lang has only the small lexicon, around 300 words https://github.com/GrammaticalFramework/GF/blob/master/lib/src/abstract/Lexicon.gf
[19:09:25] <begiak> [ GF/Lexicon.gf at master · GrammaticalFramework/GF · GitHub ]
[19:10:03] <inariksit> and replace interact with this http://www.grammaticalframework.org/doc/tutorial/gf-tutorial.html#toc150 so you don't need to start it over all the time
[19:10:07] <inariksit> (if you didn't already :)
[19:10:38] <Gurkenglas> Oh. Umm. <Gurkenglas> Is there a gigantic prebuilt pgf that I can use for a chatbot?
[19:10:42] <inariksit> I'll need to go soon, but it seems that you've got things working fine!
[19:10:56] <inariksit> hmm not that I know, and probably not now that 3.7 is very new
[19:11:06] <inariksit> sec, I'll look for a .gf to compile
[19:11:12] <Gurkenglas> gfs also work.
[19:12:06] <inariksit> try one of these? https://github.com/GrammaticalFramework/GF/tree/master/examples/translator
[19:12:07] <begiak> [ GF/examples/translator at master · GrammaticalFramework/GF · GitHub ]
[19:12:20] <inariksit> hmm ok that's a year ago
[19:12:21] <inariksit> sorry
[19:13:23] <Gurkenglas> Maye I should just rollback to the gf version of back then?
[19:13:30] <inariksit> https://github.com/GrammaticalFramework/GF/tree/master/lib/src/translator this seems to be the best translator now
[19:13:32] <begiak> [ GF/lib/src/translator at master · GrammaticalFramework/GF · GitHub ]
[19:14:03] <inariksit> no no, it's just compiling it :) and then prepare for having simple sentences with 100000 parses
[19:14:10] <inariksit> where each possible word is a string literal
[19:14:20] <inariksit> I really don't know how they make this so nice :-D http://cloud.grammaticalframework.org/wc.html
[19:14:21] <begiak> [ Demo: GF Wide Coverage Translation ]
[19:14:24] <Gurkenglas> (In the end, I don't even need to translate. Its main function will be processing something like textbooks.)
[19:14:33] <inariksit> okay
[19:15:01] <inariksit> in the summer school there's going to be one lecture about the secrets of pre- and postprocessing, how a GF grammar turns into the nice app http://cloud.grammaticalframework.org/wc.html
[19:15:01] <begiak> [ Demo: GF Wide Coverage Translation ]
[19:16:03] <inariksit> because honestly, if you parse "I am old" in the translator grammar, it's gonna give so much crap, like "maybe you meant "old" as a string literal"; and especially for common words, there are so many synonyms
[19:16:49] <Gurkenglas> I guess it should attach weightings to each possibility?
[19:16:55] <inariksit> yeah, it does
[19:17:02] <inariksit> at least  PGF2 does
[19:17:21] <inariksit> the two APIs are kinda blurred in my head :-P one of the at least does
[19:18:10] <inariksit> are you going to need to parse realistic text?
[19:18:19] <inariksit> as in, not complete sentences, colloquial speech etc?
[19:19:50] <Gurkenglas> I hope not, though my instructor might expect it.
[19:19:55] <inariksit> okay
[19:20:13] <Gurkenglas> (Because that's about as easy as detecting pictures with a bird in them, right?)
[19:20:17] <inariksit> ^_^
[19:20:25] <inariksit> yeah
[19:21:06] <inariksit> the architecture of the translator is basically: a) the basic GF resource grammar library b) lots of new words c) new grammatical extensions, which overgenerate and produce massive ambiguity d) chunks
[19:22:21] <inariksit> so, without chunks, it's neat S -> NP VP; with chunks, it's just that the sentence is chunked, and you can combine any amount of chunks into an S
[19:22:42] <inariksit> handy if it's ungrammatical or OOV, but of course they also show up in well-formed sentences
[19:23:25] <Gurkenglas> (My task is: Given a text, test students on their reading comprehension.)
[19:23:37] <inariksit> ooh sounds nice
[19:23:58] <inariksit> any chance you could come to the GF summer school in malta? in 2 weeks or so
[19:24:22] <Gurkenglas> nah
[19:24:31] <inariksit> hihi, just tried :-P
[19:25:27] <inariksit> okay, my final words before I leave for today! so, if you find that the translator produces way too much crap and you just want to get started with simple unambiguous structures, you can add DictionaryEng here https://github.com/GrammaticalFramework/GF/blob/master/lib/src/english/LangEng.gf
[19:25:28] <begiak> [ GF/LangEng.gf at master · GrammaticalFramework/GF · GitHub ]
[19:25:34] <inariksit> along with LexiconEng
[19:25:36] <inariksit> or replacing it
[19:25:44] <Gurkenglas> It's the project that people must do before they're allowed to do their bachelor thesis. (Inventing words here, I'm from Germany.)
[19:25:54] <inariksit> okay
[19:26:05] <Gurkenglas> To give you a sense for how small it's supposed to be :P
[19:26:11] <inariksit> ok :)
[19:26:37] <inariksit> well, good luck, I'm off now!
[19:54:26] <begiak> apertium: jonorthwash * 61113: /languages/apertium-rus/scripts/corpconv/: rnc2cg.py, cglib.py: some verbosity stuff
[19:58:41] <vin-ivar> Gurkenglas, wie heißt es denn?
[19:58:53] <Gurkenglas> Semesterprojekt
[19:59:17] <vin-ivar> cool
[20:26:59] <begiak> apertium: unhammer * 61114: /languages/apertium-nob/apertium-nob.nob.dix: Summary: unntatt/foruten__cnjcoo (smenob)
[20:32:00] <begiak> apertium: unhammer * 61115: /languages/apertium-nob/apertium-nob.nob.dix: Summary: actually these require «når» after, so not clb
[20:34:29] <begiak> apertium: unhammer * 61116: /trunk/apertium-sme-nob/apertium-sme-nob.sme-nob.dix: earret→unntatt/forutan
[20:36:59] <begiak> apertium: unhammer * 61117: /nursery/apertium-sme-sma/modes.xml: lex-sel modes named lex for less rsi
[20:42:03] <begiak> apertium: unhammer * 61118: /trunk/apertium-sme-nob/apertium-sme-nob.sme-nob.dix: great example of the dangers of dynamic compounding («dyregodtforhold»)
[20:44:30] <begiak> apertium: unhammer * 61119: /nursery/apertium-sme-sma/t/: latest-regression.results, latest-pending.results: t
[21:09:33] <begiak> vislcg3: tino * 10817: /tools/vislcg3/trunk/src/GrammarApplicator_runRules.cpp: Fix segfault