[01:07:33] do we have bots as contributors? 🤔 [01:07:53] gf-mode is now on melpa: http://melpa.org/#/gf [12:09:24] *** Joins: drbean (~drbean@TC210-63-209-214.static.apol.com.tw) [12:09:59] *** Quits: drbean_ (~drbean@TC210-63-209-27.static.apol.com.tw) (Read error: Connection reset by peer) [12:26:32] woo! [12:27:14] grr found another bug in gftest [12:27:37] someone should write a tool that automatically and systematically tests gftest [12:27:39] :-P [12:56:10] inariksit: do you have unit tests or quickcheck properties? [12:57:12] nope [12:57:15] :--D [12:57:21] come on who does that [12:57:23] 8D [12:57:24] okay [12:57:28] I probably should [12:58:51] I forgot the stupid hash in the GF grammar, then I ran gftest -o -c N, and no differences; but turns out that gftest at some point stopped testing functions of arity 0 O_o [12:59:14] well, "at some point stopped" = I did something stupid [14:43:42] okay I found the bug [14:44:00] I mean, fixed the bug [14:44:08] found its existence 2 hours ago [14:44:32] and in fact it comes from a piece of code that koen wrote originally :-D [14:44:49] well, still unfortunate to not have noticed it before :-P [14:47:50] also I got rid of one awkward function and wrote a much simpler one-liner; and made it possible to add a custom start category for -o mode [14:47:54] so not totally wasted 2 hours [15:14:05] that error in the bug, matching c@dupCons instead of c@#dupCons, gave some pretty hilarious morphology [15:14:22] "error in the bug" wow I am really good at expressing myself today :-D [15:17:05] it was supposed to match the last vowel before a duplicated consonant, and now it matches the first vowel and binds the rest of the word in a variable called dupCons [15:17:10] so getrouwdtrouwde [16:13:46] *** Quits: odanoburu (uid302941@gateway/web/irccloud.com/x-guduoeoqecbruhel) (Quit: Connection closed for inactivity) [17:49:20] *** Joins: odanoburu (uid302941@gateway/web/irccloud.com/x-ijcrbudliwlwlntq) [17:50:53] I found out yesterday that Koen is the author of quickcheck! [17:51:33] I still don't use it in my projects though T.T (as I should..) [17:59:31] no shame [17:59:37] I don't use it either that much, and he's my supervisor [18:00:42] I know, but it's just perfect to check a parser/printer combo! [18:01:13] :-P [18:01:31] I have on the other hand used quickcheck for totally useless stuff, like this randomisation of words and orders was created by quickcheck https://listenmaa.fi/noms-en-codi/ [18:01:39] (cause of course you write haskell code to produce html!) [18:01:50] also creating nonsensical constraint grammar rules, check [18:01:52] or automata [18:03:51] so colourful! [18:04:15] I object the notion of uselessness [18:04:23] if it's fun it's not useless [18:04:27] ^^