00:01:37 03gammafunk02 07* 0.21-a0-357-g1680097: Reduce average number of levels needed for unique Pan lords (hellmonk) 10(22 minutes ago, 1 file, 3+ 3-) 13https://github.com/crawl/crawl/commit/1680097d8593 00:19:29 03advil02 07* 0.21-a0-358-g8928369: More constraints on partial species name in options (|amethyst) 10(4 minutes ago, 3 files, 6+ 5-) 13https://github.com/crawl/crawl/commit/8928369ccc27 01:09:24 Unstable branch on crawl.jorgrun.rocks updated to: 0.21-a0-358-g8928369 (34) 03:21:53 Unstable branch on crawl.beRotato.org updated to: 0.21-a0-358-g8928369 (34) 05:07:59 Hi; I'm getting a crash on tiles 0.20.1 when I resize the window to around zero width 05:08:53 The fix is trivial; should I just send a PR to crawl/crawl on github? or does it need an issue first 10:48:16 aidanh: I often do both and link the PR from the mantis issue 10:48:28 but I don't believe that's necessary 10:48:37 aidanh: just a PR is fine, we use mantis for issue tracking (https://crawl.develz.org/mantis/), but if you have a solution already feel free to go ahead and submit it! 10:49:26 johnstein, advil, thanks 10:50:10 turned out to be more my wm's fault (ignoring minimum size hints) 10:51:40 advil: I noticed that the tiles regions code seems very tile-oriented; e.g. things like the sidebar stats and tabs are tile-aligned 10:52:38 I was interested in making the sidebar fixed-width and stick to the right, and display partial tiles in the dungeon map view to fill the space 10:52:52 Would that sort of thing be mergeable? 10:57:40 that sounds like a reasonable thing to do. But I'll warn you that one of the annoying things about deailing with local tiles is that everything has to work both for local tiles and console, which makes changes like this more complicated than they seem 10:57:58 I suspect by tiles-oriented you actually mean grid-oriented, where the grid is basically the console text grid 10:58:48 btw ideally crawl shouldn't crash even if the WM is bad (though I know there are crashes on too-small console windows as well) 10:59:35 I mean tilereg-* 11:00:08 Which has dx, dy, mx, my members, as well as a set of pixel offsets 11:00:45 Different subclasses seem to have a different grid size (e.g. monospace text) 11:02:04 But as far as I can see, these are only used in a tiles build 11:02:07 ah ok. the HUD for local tiles/console is mostly generated using cgotoxy in output.cc 11:02:20 yes, those are tiles-specific 11:04:23 -!- yesno_ is now known as yesno 11:05:19 somewhere there was an implementable for the small window crash, but I can't find it right now 11:05:41 oh yeah, https://crawl.develz.org/mantis/view.php?id=6498 11:05:49 I got a few stack traces when reducing window width, but they were all different 11:05:56 that one is console-specific but some principle 11:06:54 *same 11:08:00 Right; I'll see what I can do with wrangling the UI first, I think 11:08:34 any contributions are welcome :) 11:08:41 we don't have enough people working on local tiles 11:12:24 :) I'll see what I can do 11:33:32 03advil02 07* 0.21-a0-359-gb7bfbb7: Clarify berserk messaging (0011256) 10(2 minutes ago, 1 file, 1+ 1-) 13https://github.com/crawl/crawl/commit/b7bfbb70b25c 11:33:42 !source _do_berserk_no_combat_penalty 11:33:42 1/1. https://github.com/crawl/crawl/blob/master/crawl-ref/source/main.cc#L204 11:34:10 somewhat confused by that function, as the subsiding rate is steep enough that that third message never seems to get triggered? 11:35:15 if so they could just be collapsed, with only the "nearly" message remaining, but am I missing something about how that rather baroque penalty system works? 11:41:24 johnstein: when you have a moment, could you add 'adviltest' as an admin on cbro so I can use wizmode on that account? not urgent 12:09:22 03advil02 07* 0.21-a0-360-g65c131b: Fix boosted HP/Str coloring for berserk cooldown in webtiles (11257) 10(5 minutes ago, 1 file, 2+ 2-) 13https://github.com/crawl/crawl/commit/65c131be345f 12:09:31 Unstable branch on crawl.jorgrun.rocks updated to: 0.21-a0-359-gb7bfbb7 (34) 12:52:51 <|amethyst> advil: I think I prefer "begin to subside" to "nearly subside", since that better implies that it definitely will be subsiding soon 12:53:40 I had "begin to subside" at first, but the thing is that that message never comes >1 turn before the end, and you sometimes see it on the same turn as berserk ends 12:53:50 <|amethyst> ah 12:53:50 "begin to subside" makes it seem like you might have more time than that 12:54:01 that's why I was puzzled about that penalty code 12:54:38 you do very rarely see the later penalty message instead, but never both in the same berserk 12:59:14 <|amethyst> hm 12:59:30 <|amethyst> I guess you'd pretty much have to have duration extension to see that 12:59:46 yeah 13:00:01 I've seen "blood rage quickly leaving you" sometimes 13:00:01 <|amethyst> since according to the table above that function, you've already lost 20 turns off duration by the time the penalty gets up to 5 13:00:40 <|amethyst> and the starting duration for berserk is no more than 19 turns 13:09:19 Unstable branch on crawl.jorgrun.rocks updated to: 0.21-a0-360-g65c131b (34) 13:24:20 Unstable branch on crawl.akrasiac.org updated to: 0.21-a0-360-g65c131b (34) 13:34:08 haha 13:34:09 d - tha Incunabulum o' Evil Good Luck 13:59:12 New branch created: pull/620 (2 commits) 13https://github.com/crawl/crawl/pull/620 13:59:12 03Aidan Holm02 07https://github.com/crawl/crawl/pull/620 * 0.21-a0-361-g87231db: Add GL render clipping support for local tiles 10(3 hours ago, 3 files, 18+ 0-) 13https://github.com/crawl/crawl/commit/87231dbf6007 13:59:12 03Aidan Holm02 07https://github.com/crawl/crawl/pull/620 * 0.21-a0-362-g81f2820: Improve layout+spacing of tiles main dungeon view 10(41 minutes ago, 2 files, 22+ 6-) 13https://github.com/crawl/crawl/commit/81f282034a8e 14:03:46 advil: well, the tilereg code is pretty gnarly, but that's a start 14:04:29 I remember seeing that unifying menu implementations was a goal; are there any similar goals with the tiles UI stuff? 16:46:50 so I just won a game on CDO and the highscores there are... interesting: https://bpaste.net/show/40c8c71a0c48 16:48:36 I don't know whether this is just some issue with the relevant file on CDO getting messed up or something more general, but clearly something interesting is going on 16:49:06 I think all the score stuff is waiting a full rebuild? 16:49:13 (db rebuild) 16:50:46 geekosaur: this isn't the CAO scoring pages 16:50:58 no, but I thought cdo had issues too 16:51:04 ah 16:51:56 also the snark said something about needing to take down sequell at some point and rebuild its databases too. general corruption or confusion somewhere, I think? 18:14:22 Unstable branch on underhound.eu updated to: 0.21-a0-360-g65c131b (34) 18:45:10 The build has errored. (master - b7bfbb7 #8804 : advil): https://travis-ci.org/crawl/crawl/builds/287106102 19:11:55 advil: i just read your reddit post and i don't see why you'd have to pregenerate everything. if we had the ability to produce a single floor from a seed, you don't need to produce the whole dungeon all at once: just generate the player's current floor and choose seeds to be used for each other floor they gained access to 19:12:30 ie, after generating D:1, save the seed for D:2, to be used when they go downstairs. after generating D:8, save seeds for both D:9 and L:1 19:15:02 are you accounting for level visitation order? 19:15:17 i.e. someone gets shafted 19:15:32 amalloy: ^ 19:15:39 I may be coming into this discussion without enough context 19:16:04 but if it was about making the dungeon generate the same way from a seed, there are some subtle issues 19:16:08 for one, consider vaults 19:16:41 vaults currently have a mechanic where they cannot place more than one time without a special tag 19:16:55 gammafunk: i was attempting to account for level visitation order by saving seeds for each floor you could exit to, yes. that doesn't work for shafts but it's a pretty small change to incorporate shafts as well 19:17:26 vaults are still an issue though 19:17:29 yes 19:17:30 eg, if the player does get shafted you can spend a little extra time generating the levels they fell "through", or you can predetermine shaft targets and treat them as regular exits 19:17:38 i don't get why vaults are an issue 19:17:47 I guess generating the levels that you fall through works 19:18:04 yeah you could do that maybe 19:18:06 well, not completely 19:18:26 if you have a vault that places both in D and in lair and can't place multiple times 19:19:22 then I don't see a good way of handling that unless we can actually pre-generate the list of vaults in every level at the start of the game without actually making the levels 19:19:44 yeah, which isn't possible given how we generate levels 19:19:57 since it involves making it and seeing if the level is coherent, and if not doing another selection 19:20:21 or in other words you can't know if choices for a level are valid until after it's fully generated 19:20:22 ah, right. floors aren't independent 19:21:19 imo remove vaults 19:21:29 m-my v-vaults....;_; 19:22:17 it would be nice if we had level gen were you could somehow say "I want this layout and these vaults, make it work!" 19:22:49 I mean, you can sort of do that already, which is what &P does, but you can't make any reasonable guarantees like that 19:31:15 gammafunk: so after winning my gnoll and looking at some dumps, the main thing I'm unhappy about is that they get too good at spellcasting in late game 19:31:39 so I'm thinking of reducing their spell school apts to +6 19:39:46 elliptic: you feel that the early game is alright with how the starting XP is spread? 19:40:04 Some where complaining about that, but I mean it's part of the species idea that early levels are easier 19:40:27 I mean, I don't think the early game stays easy for that long 19:41:12 after D:3 or so your relevant skills are probably lower than they would be as a non-gnoll 19:41:28 yeah 19:41:32 you just have much more flexibility to switch gear/spells depending on what you find 19:41:47 well decreasing magic apts to +6 seems reasonable 19:41:55 would you leave spellcasting itself at +8? 19:42:18 the spellcasting skill itself, I mean 19:42:29 yeah, I was thinking just change the individual schools 19:42:44 spellcasting being high means more spell slots, which is nice 19:43:00 right 19:43:15 alright, I can make that change, or if you feel like doing it, feel free 19:43:25 I was also going to change invo apt to +9, primarily for title purposes 19:44:10 seems fine, yeah 19:44:31 none of these apt changes really have a large impact, it seems like +-1 to an apt translates to about +-1 to the skill level at most points 19:45:03 -!- ProzacElf_ is now known as ProzacElf 19:46:59 gammafunk: is your impression that people generally find gnolls on the easy side of things? I did some queries and its winrate/runerate has been on the high side of things but not absurd 19:47:20 elliptic: there have definitely been a lot of people saying that, yeah 19:47:35 (sample is of course somewhat low and biased by things like people trying the new species) 19:47:41 someone on tavern suggested that we halve gnoll scores! 19:47:46 which of course is an outstanding idea 19:48:32 are they really that good for speedrunning :P 19:48:34 !hs * gn 19:48:35 19929. Yermak the Wrestler (L27 GnWn of Cheibriados), escaped with the Orb and 15 runes on 2017-10-10 10:57:53, with 58594819 points after 24599 turns and 7:37:11. 19:48:41 okay not bad 19:48:41 dang 19:48:46 !hs * gn -log 19:48:47 19929. Yermak, XL27 GnWn, T:24599: http://crawl.xtahua.com/crawl/morgue/Yermak/morgue-Yermak-20171010-105753.txt 19:50:01 20 in all skills 19:51:11 I did note that gnolls are pretty good at getting 27 in all skills compared with other species, unsurprisingly 19:51:17 but still seems to take a couple zigs 19:51:50 someone reported all 27 after one zig, if I read their post correctly 19:51:56 but I'm not sure what else they had done 19:52:06 that sounds possible, yeah 19:52:25 !lg Majang gnfi zigscompleted>0 19:52:26 No games for Majang (gnfi zigscompleted>0). 19:52:40 !lg Majang gnfi 19:52:42 No games for Majang (gnfi). 19:52:47 I'm not sure whether it is a problem that gnolls become ridiculous with enough xp - I was vaguely considering the idea of capping their skills at 20 or some such 19:52:55 https://crawl.develz.org/tavern/viewtopic.php?p=322072#p322072 is the post fwiw 19:53:02 or halving their skill xp at some point 19:53:16 but maybe not worth worrying about quite yet 19:54:03 right, I suppose a cap wouldn't really harm anything, but it's kind of bad in how it affects different schools in different ways 19:54:37 I wonder if you could base an aptitude decreas on XL 19:55:49 basing it on total xp used on skill training was what I had in mind for decreasing apts at some point 19:57:25 you wouldn't want to simply decrease apts though since you wouldn't want skills to suddenly decrease 19:57:47 really you'd want to increase skill point cost going forward 19:57:57 oh, right 19:58:05 would be easier with some of the skill training rewrite stuff that was discussed a while ago 19:58:43 anyway, it is an option if we want to weaken lategame gnolls without touching earlier parts 20:03:55 hm, I'm looking at gn starting skills currently and I guess they are a bit higher than I remembered, maybe nerfing them slightly isn't a bad idea 20:04:23 I do like that your background remains relevant for much longer than normal though 20:05:29 after magic goes to +6, it will be a bit different though 20:05:42 for backgrounds that use those skills, that is 20:06:25 I guess some backgrounds also just have higher starting skill levels 20:06:54 I checked Cj first and was a little taken aback by 7.0 conj skill even with +6 apt 20:07:27 but other magic backgrounds seem more reasonable 20:09:13 I'll leave that alone for now and just do the apt changes 20:15:43 I would lean towards increasing skill cost or decreasing skill xp gain over a hard cap if it is needed 20:16:40 yes 20:16:44 mainly from my experience with messing with hard capping from an earlier version and the feedback from it 20:17:00 03elliptic02 07* 0.21-a0-361-g4efd3f9: Adjust Gnoll apts: spell schools to +6, Invocations to +9. 10(7 minutes ago, 1 file, 13+ 13-) 13https://github.com/crawl/crawl/commit/4efd3f98658c 20:18:08 well people like caps, they fit on your head and you can enchant them and maybe they have a good ego 20:18:11 just saying 20:18:26 those are hats, not caps 20:18:35 caps don't exist anymore 20:19:06 hats are totally just caps, you're not fooling me with that nonsense at all 20:38:36 amalloy: yeah, there's definitely got to be a way to do it by pregenerating some future seeds from a seed or something. All I meant there though was that (IIRC) bel pregenerated everything 20:51:42 the vault issue that we talked about is not easy to solve 20:52:13 there would have to be some fundamental changes to vault placement 20:57:59 that does sound hard 21:09:27 Unstable branch on crawl.jorgrun.rocks updated to: 0.21-a0-361-g4efd3f9 (34) 22:36:13 -!- axon is now known as Guest25172