00:12:54 The build was fixed. (master - 71aafec #11203 : advil): https://travis-ci.org/crawl/crawl/builds/502907039 00:24:36 Unstable branch on CRAWL.XTAHUA.COM updated to: 0.24-a0-169-g71aafec (34) 01:00:40 Stable (0.22) branch on CRAWL.XTAHUA.COM updated to: 0.22.1-57-g7394efd 02:54:52 -!- CaptainFruitcake is now known as CanOfWorms 03:40:19 -!- amalloy is now known as amalloy_ 03:51:11 Stable (0.22) branch on crawl.kelbi.org updated to: 0.22.1-57-g7394efd48c 04:13:23 Fork (bcadrencrawl) on crawl.kelbi.org updated to: 0.22.1-76-gb49813fe60 09:15:37 heh in a debug DGL build you are in wizard mode but the & key appears to be disabled 09:23:48 am I hallucinating that there was in the past error checking for invalid keys in rc files? 09:25:47 alexjurk1: you are using mojave right? have you ever noticed anything like https://crawl.develz.org/mantis/view.php?id=11895 in offline tiles? 09:26:00 just occurred to me it could be a mojave-specific issue, and that's why I can't replicate it 09:26:56 03advil02 07* 0.24-a0-170-gd8427f5: Add a warning for setting `game_seed` online 10(9 minutes ago, 1 file, 8+ 2-) 13https://github.com/crawl/crawl/commit/d8427f5388a2 09:32:29 Unstable branch on crawl.kelbi.org updated to: 0.24-a0-170-gd8427f5388 (34) 09:51:48 I'm sort of tempted to actually play the gotm thing myself 12:25:51 Stable (0.22) branch on crawl.akrasiac.org updated to: 0.22.1-57-g7394efd 12:30:49 Unstable branch on crawl.akrasiac.org updated to: 0.24-a0-170-gd8427f5 (34) 12:43:51 -!- amalloy_ is now known as amalloy 15:06:02 !crashlog 15:06:18 20738. guttsta, XL25 VSFi, T:41328 (milestone): https://underhound.eu/crawl/morgue/guttsta/crash-guttsta-20190305-235710.txt 15:06:49 hm. nefhilion in ##crawl is reporting their 0.23 game is failing to load 15:07:19 !lm nefhilion x=cv,src,gid 15:07:20 2727. [2019-03-07 20:01:55] [cv=0.24-a;src=cbro;game_key=nefhilion:cbro:20190207194335S] nefhilion the Changer (L7 DrTm of Jiyva) left a sewer on turn 3430. (Sewer) 16:06:29 New branch created: pull/1004 (1 commit) 13https://github.com/crawl/crawl/pull/1004 16:06:29 03NormalPerson702 07https://github.com/crawl/crawl/pull/1004 * 0.24-a0-140-g284c72d: Fix cut-off ability cost description (11893) 10(14 minutes ago, 1 file, 5+ 5-) 13https://github.com/crawl/crawl/commit/284c72dd798a 16:07:14 -!- aidanh_ is now known as aidanh 16:22:28 03NormalPerson702 07https://github.com/crawl/crawl/pull/1004 * 0.24-a0-140-g85ad28d: Fix cut-off ability cost description (11893) 10(29 minutes ago, 1 file, 5+ 5-) 13https://github.com/crawl/crawl/commit/85ad28d2663c 16:39:42 advil: i'll test now 16:49:48 advil: can't seem to replicate. AIUI the bug appears if you have a spell menu with two spells? 16:50:17 https://imgur.com/a/2rAZ86B 18:22:49 alexjurk1 ok thanks; I haven't been able to replicate it either 18:23:00 Unstable branch on underhound.eu updated to: 0.24-a0-170-gd8427f5388 (34) 20:26:46 -!- amalloy is now known as amalloy_ 20:55:25 advil: seed bug with morgues if you check out the top games in the weekly challenge leaderboard, the morgue files sometimes show '0' as the seed https://crawl.project357.org/ 20:56:11 you can see all three games have an identical dungeon though 20:56:49 ebering: it looks to me like there's a small mistake in the crawl lua documentation on your website, and you.contaminated returns a number and not a boolean. (also the comment in l-you.cc says it's returning a boolean, which I imagine is where the mistake came from. I haven't dug into how l-you.cc works, but it looks like the lua function uses get_contamination_level in player.cc which returns an int) 20:58:20 glaas: AFAIK the documentation is auto-generated from the comments in l-you.cc, so a pull request to update them should fix the issue 20:58:53 aha. all right, in that case i will fix it myself a little bit 21:01:59 advil: i suspect on the basis of absolutely nothing that if you play your game in one shot, the seed shows up correctly 21:02:22 actually, that's definitely not true. I'm clueless 22:53:59 -!- engineering is now known as bitchbutt 23:01:02 %git 3f7990ae4ddb0e2 23:01:02 07advil02 * 0.24-a0-161-g3f7990a: Save game seed properly 10(3 days ago, 1 file, 1+ 0-) 13https://github.com/crawl/crawl/commit/3f7990ae4ddb 23:01:05 alexjurk1 ^ 23:07:11 ooh, cool 23:07:40 time travel bug fixing 23:08:00 I'm very skilled 23:08:13 (it's one of the things that popped up in uv4's gotm) 23:09:26 advil: I was wondering, are there any general strategies for seed searching? As in I want to make make a "difficult" seed according to some criteria 23:09:52 I could certainly just generate a bunch of dungeon and do something akin to objstat, but track dungeons individually 23:09:56 not yet! 23:10:09 yeah, I do plan on producing a seed catalog for a few hundred seeds 23:10:16 but finding something that meets more than one criteria could be a challenge 23:10:23 yeah 23:10:30 I'm not sure how to do it beyond brute force 23:10:48 I guess a general strat is something like, generate N seeds matching criteria 1, for a reasonable, N, then search through those to find ones matching criteria 2, ... 23:12:05 except I think the first step is really generate N seeds 23:12:20 though for very early decisions it might be more manageable 23:12:26 e.g. temple layouts 23:12:31 maybe just easier to generate N seeds and make a catalogue of everthing in them? 23:12:35 yeah 23:12:59 that's what I'm going to do at least 23:13:06 apparently the seed uv4 randomly picked happens to have a corrupted temple, which is kind of fun 23:13:30 that is fun, but I think the best thing overall was ebering's zot:1 arrival vault 23:13:44 heh I haven't looked at that yet, except I saw some comments about it 23:13:52 his "three ring circus" vault and you arrive in the center ring surrounded by like 8 tmons 23:14:01 nice 23:15:44 if our use of rng was more disciplined in various was it'd be possible to do some search, pcg supports some fairly fancy fast forward techniques 23:16:05 but as it is, the connection between the raw numbers and the actual results is *incredibly* complicated 23:16:17 *various ways 23:27:45 hrm, I guess one thing it might be easy to do is to hook up pregen to objstat 23:28:57 really wish we had a kind of "libcrawl" that could be used to make a tool to just generate levels and save them and then provide access for a tool that could read them 23:29:41 maybe just running objstat on one iteration with a seed would be useful though 23:47:42 generating a thousand pregen saqve games should be pretty easy with an expect script 23:48:03 the hard part is probably taking a save game and figuring out "does this meet my criteria" 23:52:41 well that's what I'm talking about; a tool that can read saves 23:52:49 objstat is one approach but it's coded in-game 23:52:54 which is just not ideal 23:59:51 Unstable branch on CRAWL.XTAHUA.COM updated to: 0.24-a0-170-gd8427f5 (34)