00:00:52 Unstable branch on CRAWL.XTAHUA.COM updated to: 0.24-a0-106-gcea5463 (34) 01:25:22 -!- amalloy_ is now known as amalloy 01:28:27 Unstable branch on crawl.develz.org updated to: 0.24-a0-106-gcea5463 (34) 01:37:30 Allied shambling mangrove's roots only target itself 13https://crawl.develz.org/mantis/view.php?id=11877 by Tmoiy 03:00:08 Monster database of master branch on crawl.develz.org updated to: 0.24-a0-106-gcea5463 03:30:58 Fork (bcrawl) on crawl.kelbi.org updated to: 0.23-a0-1682-g99b902e015 03:56:32 -!- amalloy is now known as amalloy_ 04:11:34 Can see item that's not in LOS where I have not explored before 13https://crawl.develz.org/mantis/view.php?id=11878 by Flugkiller 07:25:45 -!- amalloy_ is now known as amalloy 08:57:10 Hey! 08:58:26 I want to try my hand at fixing https://crawl.develz.org/mantis/view.php?id=11833 08:59:42 But to be honest, I'm lost regarding as to which part of the code I should look at, the bug is really weird, it only happens in the right and bottom of the screen, and on tiles. 09:00:09 Does anyone have any idea about what part of the code is tasked with drawing the targeting overlays in tiles? 09:00:50 I was thinking maybe tile_place_ray, but it does not seem to be the case 09:31:53 Nasst- I haven't really looked into that bug but that seems a plausible place to start; tile_place_ray does not draw the ray itself, but the calling code does. You might look at direction_chooser::draw_beam_if_needed 09:32:41 and viewwindow 09:35:23 man, that code is pretty dodgy 09:37:05 my immediate guess would be that something is going wrong with the hacky global num_tile_rays that is apparently used to determine how far to draw 09:41:43 !tell gammafunk something is weird with the tournament streak leaderboard, yermak is listed with 4 different ongoing streaks 09:41:44 advil: OK, I'll let gammafunk know. 09:42:45 could be related to having multiple live games on different servers? 09:46:54 I'm not sure this is a bug: the definition of streak in the tournament rules says explicitly that you can have multiple active streaks 09:47:15 and it seems that the multiple yermak streaks all fit the definition 10:20:50 Thenks advil! 10:20:59 I'll look into it. 10:21:27 By the way, I submitted a patch for this bug: https://crawl.develz.org/mantis/view.php?id=11860 10:21:40 Should I do a PR on github instead? 10:38:59 I saw that, a PR is easier for us to keep track of 10:46:20 advil: not sure if you saw the response, but that listing is just a byproduct of how the tournament counts streaks 10:46:48 basically a game started after all games in independent streaks will count for all of them 10:47:04 so that listing isn't wrong per se 11:32:53 oh, I forgot that it counted streaks that way 11:32:53 so all of those involve some complicated concurrency situation 11:44:28 oh nice, I think both of the more mysterious lingering seed bugs were due to acquirement-based item generation 12:06:26 Unstable branch on crawl.akrasiac.org updated to: 0.24-a0-106-gcea5463 (34) 12:34:02 ebering: did you have a fix lined up for the unique placement bug with multiple DEPTHs? 12:50:13 looking at this my instinct is to make it a validation error if there is more than one DEPTH line, currently that catches exactly this bug and nothing else 13:12:34 I was wondering about that, is there any info on that bug? 13:12:48 is it causing unique generation to be weird? 13:13:49 03David Damerell02 07https://github.com/crawl/crawl/pull/979 * 0.24-a0-69-g8a4706e: checkwhite / unbrace 10(9 minutes ago, 2 files, 285+ 286-) 13https://github.com/crawl/crawl/commit/8a4706e9ec69 13:19:16 yes 13:21:06 %git 402804a59a51 13:21:07 07gammafunk02 * 0.22-a0-899-g402804a: Cleanups for uniques.des (amalloy) 10(7 months ago, 1 file, 10+ 5-) 13https://github.com/crawl/crawl/commit/402804a59a51 13:21:30 all the uniques touched in that commit are generating only for the depth range on the second line, which is usually pretty narrow 13:22:04 e.g. aizul only in crypt 13:22:08 crypt:1 even 13:22:20 I wonder if this has affected the kill-all-uniques race? 13:22:46 haha yeah 13:23:22 yermak is missing both aizul and jorgrun for example 13:23:27 aha 13:23:41 dynast is missing several of those as well 13:24:08 hrm 13:24:30 but I've definitely fought asterion outside of crypt 13:24:30 !lm . uniq=asterion 13:24:36 or have I 13:24:45 for some of them, there are a few vaults that will specifically place them 13:24:45 oh I see 13:24:50 misread the diff 13:25:10 oh yeah asterion isn't affected 13:25:11 a simple fix is to mostly revert that commit 13:25:21 I only did that to avoid long line lengths 13:25:37 yeah, I can do it later, I would add this validation check at the same time 13:25:40 really should have considered if Depth works that way when I made the commit, oh well 13:25:48 45. [2018-09-01 22:57:55] gammafunk the Spear-Bearer (L18 OpGl of Nemelex Xobeh) killed Asterion on turn 44421. (Elf:1) 13:25:51 was just checking in case ebering started working on it, since he's the one that found the problem 13:25:54 it's not documented 13:26:02 I will add documentation at the same time as I add the check 13:26:23 So I made some progress on the Barachim wand of digging bug 13:28:26 In particular, changing '#define LOS_RADIUS 8' to 9 solves the issue without introducing any new issues, apparently. So whatever the problem is it probably depends on LOS_RADIUS 13:29:14 I'm going to try changing LOS_MAX_RANGE independently of LOS_RADIUS to see if the problem can be narrowed down further 13:30:27 hrm 13:30:43 if that depth is being completely overriden 13:30:46 !lm . uniq=jorgrun 13:31:19 I'm not aware of a vault that places him 13:31:38 yeah he appears in no vaults other than uniq_jorgrun 13:31:46 !vault uniq_jorgrun 13:31:47 1/1. https://github.com/crawl/crawl/blob/master/crawl-ref/source/dat/des/builder/uniques.des#L162 13:31:55 aha 13:32:01 was later re-added to swamp 13:32:07 after that commit, so that makes sense 13:32:16 90s limit exceeded: killed !lm . uniq=jorgrun 13:32:20 yeah I found him in swamp on my current mummy 14:00:28 advil: I did not have a fix lined up 14:40:37 =nemelex GnHu 14:40:40 Defined keyword: nchoice => GhVM|TeEn|VpCj|VpAE|CeAE|DDIE|DEMo|FeWr|HaEE|GnHu 15:20:43 ebering: I see what you were worried about with default depths, but I can't actually find any evidence that lua dgn.default_depth is ever used as a way of setting them 15:21:48 everything uses dgn.set_default_depth, which works via an entirely different thing 15:21:58 if I go ru and take hated by all, can i still cast butterflies? 15:22:09 i have something really awesome in mind 15:22:22 No 15:22:30 even after abandonment?> 15:22:49 Since abandoning Ru doesn't undo any sacrifices... 15:23:00 it flat-out stops all summoning? 15:24:00 so what if i, say, abandn ru with ahted by all and use lungonu corruption?this isn`t my original idea or even a smart one, just want to know 15:24:28 gammafunk help me elliptic help me 15:24:31 Lucy corruption summons are neutral, not friendly; you'd have to test that in wizard mode. 15:24:38 ebering help me 15:25:09 !klownpie twelwe 15:25:09 The Killer Klown throws a pie at twelwe. The Klown pie hits twelwe! Moon pie! twelwe turns into a filthy swine. 15:25:16 :O 15:25:29 i want to Ru and rejecting everything until Hated By all comes up. Abandon Ru, do Dithmenos, then I`ll spam butterflies everywhere I go and get controlled blink for free on my formicid with shadowstep 15:25:40 twelwe: But you can't do that, as mentioned. 15:26:07 twelwe: sack of spiders with 0 evo is perhaps the tool you seek 15:26:09 there is something really wrong with this game imo then 15:26:11 not great way to do that though 15:26:12 if you try to cast butterflies, it will say, _You cannot coerce anything to answer your summons. 15:26:44 we haven't yet prevented use of evokers...unless I did in fact do that 15:26:44 memory a bit hazy there 15:26:46 i think it evokes the clouds but not elementals 15:26:57 well talking about sack of spiders 15:27:07 i mean even the ely vaults hated me 15:27:11 _You reach into the bag... You see a puff of smoke. x2; ...but nothing happens. 15:27:25 haha, that's a bug 15:27:34 it was not like insensitive or anything on my feelings level but really? the ely vault hated me too? 15:27:48 quokkas know that you betrayed their master 15:27:51 they are very loyal 15:28:00 to ru or ely> 15:28:10 ru is the quokka god 15:28:15 come on read your lore 15:28:25 i did and it got boring FAST 15:28:29 had to stop 15:31:01 i should have came here before posting my master formicid cblink idea to SA 15:31:17 a retraction post is now in order 15:31:35 twelwe: Just suggest it in the Gooncrawl thread, job done :-) 15:32:03 hey my account is an `03 so they might do it 15:32:10 how many SA jailings do you have 15:32:22 i`ll link you 15:32:47 only one is unjustified https://forums.somethingawful.com/banlist.php?userid=46042 15:33:00 corruption after abonding ru looks like it still works as normal, I assume because nothing it summons is your friend 15:33:05 *abandoning 15:33:26 i quoted a total creep about female phenom Gina Carano`s feet and it was taken as my own thoughts] 15:33:43 no bans! 15:33:50 15 years strong 15:35:19 ebering just to double check, was that what you had in mind with removing the clear() call? because I think it would actually work to just remove that (and remove dgn.default_depths) 15:37:22 would that allow DEPTH to be additive, like tags? 15:37:39 either way, would be nice to have a way to wrap long lines; I guess a simple \ would have worked though? 15:38:03 I use that sometimes with MONS, although I try to avoid that and just use additional MONS statements these days 15:38:19 twelwe: Vexingly Senor Lowtax has hidden the Gooncrawl thread, which might finally squeeze $10 out of me 15:38:42 yeah, \ worked, I was about to just do that but then I realized that I might be misunderstanding why DEPTH: was implemented like it was 15:38:44 those guys are nuts 15:38:52 i will never play gooncrawl 15:39:09 the gooncrawl thread comes and goes 15:39:16 if you wait long enough it'll reappear 15:39:35 there's one now 15:39:36 And patience _is_ a virtue. (It is not, however, a virtue I possess). 15:40:20 i have huge mixed feelings about SA. i`m a 16 year opium brewer due to it and like that alot, but never had any meaningful conversations otherwise. oh, and a 12k youtube channel 15:41:13 i really only post in TCC anymore 15:41:15 I can't figure out what the point of dgn.default_depth is, it was present in the snark's initial implementation of all this stuff but as far as I can tell, not used even then 15:41:34 advil: that's not just the default-depth DES statement? 15:41:38 those gooncrawl guys are nuts though 15:41:40 that is definitely used a bunch 15:41:54 git grep default-depth will show instances 15:42:13 it's to avoid specifying the same DEPTH lines over and over for vaults all using the same one 15:42:41 `default-depth` calls dgn.set_default_depth 15:42:44 not dgn.default_depth 15:43:41 what happens if you do hated by all, abandon ru, then worship hep? 15:43:45 interesting 15:44:58 twelwe: You can't. 15:45:30 you cant join fedhas or hep? 15:45:39 yred? 15:45:47 beough> 15:46:39 if you go hated by all then join dith, will your shadow mess with you too? 15:46:58 Your shadow isn't a real creature with Dith, surely. 15:47:07 it depends on whether you hate yourself 15:47:10 it appears sometimes 15:47:19 You can't join any god listed in _god_rejects_loveless() in religion.cc 15:47:22 it has to appear to do a bolt attach 15:48:00 hmm 15:48:22 i`m officially out of ideas for cheap shadowstep escapes with my formicid then 15:48:44 back to boring as fuck passwall 16:10:26 advil: what I had in mind was to replace the duplicate DEPTH line in uniq.des with linebreak \ 16:10:39 and not touch any implementation 16:11:34 that would work, but I wanted to at least clarify what the implementation is supposed to be 16:12:24 I think it would work to allow multiple DEPTH: lines to work the way other multiple lines work in .des 16:12:35 yeah; to my best understanding DEPTH is not supposed to be additive, but my best understanding is somewhat hazy 16:12:56 ok 16:13:42 I think it's also a question of what we'd like it to be 16:14:00 yeah...it's more consistent with other stuff if it's additive, right? 16:14:13 I wish I could answer that better, but TAGS is additive 16:14:34 not sure if PLACE is 16:14:53 it would probably be best if it were additive since it's purely a within-vault statement 16:14:59 I don't see a lot of application for it not being additive, but you can still in principle use lua calls to override it 16:22:09 03advil02 07* 0.24-a0-107-g431b8d2: Change behavior of multiple DEPTH: lines, fixing some uniques (ebering) 10(6 minutes ago, 2 files, 3+ 7-) 13https://github.com/crawl/crawl/commit/431b8d2fe338 16:23:16 my best guess is that ds started to implement default depths in one way, then switched gears and didn't remove the first version fully 16:30:48 there are apparently other los-related autotravel bugs showing up for clear door vaults 16:30:59 yeah, I've noticed some oddities 16:31:28 in one case o kept taking me towards the door rather than declaring it inaccessible 16:31:54 so probably there is some check for regular runed doors that isn't being done for clear ones 16:32:41 ah, yes, that's just an aspect of autotravel wrt unseen tiles 16:33:00 Unstable branch on crawl.kelbi.org updated to: 0.24-a0-107-g431b8d2fe3 (34) 16:33:00 if autotravel sees that unseen areas are behind that door, it will take you there 16:33:19 that's fairly fundamental to how autotravel works but it could be changed certainly 16:33:49 since opaque runed doors also have behaviour like that 16:34:01 if you place e.g. door vault and don't map the level 16:34:13 o will take you adjacent to the door 16:34:35 and that one uses opaque doors 16:34:57 it might be nice to change that behaviour because I think people would generally perfer a message 16:35:03 s/perfer/prefer/ 16:35:19 but it also sort of convenient I guess 16:36:30 ah 16:36:52 I thought completely opaque door vaults wouldn't do that? 16:39:25 no, they do since the same principle is at work 16:39:50 autoexplore looks for unseen areas not considering runed doors (of any opacity) as traversible 16:40:07 then does a fallback pathfind considering said doors traversable 16:40:28 if fallback pathfind can reach an unseen square, it starts travel to said square 16:40:56 and this travel just continues until the player can't move (with a special messaged when the player is stopped adjacent to a runed door) 16:42:04 for greedy items as the target when no unseen squares exist, it works a bit differently 16:42:23 with the end result that travel is not even attempted 16:42:44 so if you have a fully visible runed door vault (with clear doors, say) and items inside 16:43:04 if there are no unseeable squares in the vault, just items, you'll get a message with no travel attempt 16:43:38 but for the unseen areas inside of runed door vaults, it actually starts travel 16:45:08 actually when I said there's a special message when the player is stopped by the runed door, that's not correct; being adjacent to runed door so that this door is the next travel move is actually just the travel stopping condition 16:45:33 travel just proceeds until it sees that the next suggested move is a runed door, at which point it just stops 16:48:35 !lg niggur 16:48:35 No games for niggur. 16:48:57 ^ this is an offensive name, right? 16:48:57 someone signed up with it on CPO 16:50:36 take them out 16:51:38 yes, I'd think so 16:52:49 ok, i'll disable the account 16:53:22 man, they're a precious newbie and not responding to my chat messages though. Hope it doesn't dishearten them 16:54:39 gammafunk now I'm getting extra confused because I can't replicate the behavior I saw in my game, but you're right that that happened before. I *think* what is different is at least partly the messaging: in previous versions it would say something about a runed door, but I can't get it to do that now 16:55:12 when I've tried to replicate this in wizmode on master, I'm just getting "Partly explored, can't reach some items and places." with no movement 16:59:34 advil: but is this with a clear runed door involved or an opaque one 16:59:46 with clear ones 17:00:16 I didn't actually try in trunk with an opaque one, did test it in 0.21 to refresh my memory 17:00:43 I think the only difference is that the doors are clear, but the behaviour is actually unchanged 17:01:06 what was happening in 0.21 is different than anything I have got it to do now 17:01:21 what did you see in 0.21 specifically? 17:01:22 but it wasn't a very controlled comparison, and it isn't doing what happened in my game either 17:01:50 when I place door vault in trunk 17:01:50 in 0.21, it moved you towards the door and printed a message that was approx, "level partly explored; runed door" (can't remember the exact wording) 17:01:57 yes 17:02:00 and I see this in trunk 17:02:04 just now I tried with door_vault 17:02:13 did not map the floor first, mind you 17:02:20 just placed it, &G to clear monsters 17:02:21 then autoexplored 17:02:23 and I get 17:02:27 _Partly explored, unopened runed door. 17:02:34 as travel stops next to door vault 17:02:46 door vault being opaque, right? 17:02:48 right 17:03:06 yeah, what I mean is that I have not provoked that behavior for clear runed doors 17:03:22 yeah, and I'm saying this isn't actually due to a change to travel code 17:03:32 it's due to how travel works and since those doors are clear 17:03:43 what's a vault with clear doors and large internal areas? 17:03:48 btw you may want to update the level docs at some point to reflect the new meaning of = 17:03:56 yep 17:04:07 I was using minmay_acid_items 17:04:12 it has annoying jellies tho 17:04:20 slimes 17:04:53 yeah I just do &G and remove level monsters 17:04:56 I also tried mikee_hates_you 17:04:59 so that vault fully reveals itself 17:05:12 partly explored, can't reach some items 17:05:19 ah, I see what you mean 17:05:26 greedy items have different travel logic than unseen squares 17:05:29 so it needs to be a vault with some remaining hidden areas 17:05:31 yep 17:05:44 if a vault has unseen squares, you'll travel to it, stopping at the door 17:07:03 it kind of feels that all of these reachable things should have unified behaviour in terms of what happens when you hit 'o' 17:07:13 one of the downsides to using = for this glyph is that it's very hard to search for 17:07:26 yeah, it's kind of just an unfortunate glyph choice 17:39:06 The build passed. (master - 431b8d2 #11129 : advil): https://travis-ci.org/crawl/crawl/builds/492921019 18:22:50 -!- Tiobot is now known as Guest54956 18:23:21 Unstable branch on underhound.eu updated to: 0.24-a0-107-g431b8d2fe3 (34) 18:36:11 03NormalPerson702 {amalloy} 07* 0.24-a0-108-g011bc4b: Fix yellow draconians' rAcid mutation not being applied (malacante) 10(3 days ago, 1 file, 1+ 0-) 13https://github.com/crawl/crawl/commit/011bc4bd618b 18:47:36 Unstable branch on crawl.kelbi.org updated to: 0.24-a0-108-g011bc4bd61 (34) 18:52:42 has anyone written a morgue file parser? 18:53:00 I want to get some basic info out of morgues, like score and whether they won 18:53:12 wizmode has such a thing 18:53:17 it's probably out of date 18:53:31 wizmode is one, and also i know someone doing statistics gathering implemented something in python(?) 18:54:04 google search for `site:reddit.com/r/dcss python morgue` yields several promising results 18:54:59 (sorry to offer an offhanded comment and then leave, but i gotta go) 18:56:26 you have to get it out of a morgue? 18:56:31 those things are in the logfile 19:00:19 that's true. Even the seed is in logfile now 19:42:43 is there an option to not create *.lst and *.map files in morgue directory? 19:43:31 -!- amalloy is now known as amalloy_ 19:44:19 I was thinking about a PR to disable both files by default for online games. Does anyone use them? 19:44:32 they feel like more debugging oriented outputs 20:03:13 alexjurkiewicz: lst files I definitely use 20:03:32 although I think the UI for it could be much better, it's greatly helpful when telling people about stuff they've missed 20:03:37 since it's the stash for their entire game 20:03:52 so you can look at that and say "here's this weapon you could get on D:12 shop" etc 20:04:08 the map file is perhaps less useful 20:05:22 =nemelex KoAK 20:05:25 Defined keyword: nchoice => GhVM|TeEn|VpCj|VpAE|CeAE|DDIE|DEMo|FeWr|HaEE|GnHu|koak 20:58:57 lower case... 21:07:05 weird 21:07:07 why did it do that 21:07:11 oh 21:07:17 probably someone had already added it 21:07:31 !kw nchoice GhVM|TeEn|VpCj|VpAE|CeAE|DDIE|DEMo|FeWr|HaEE|GnHu|KoAK 21:07:32 Defined keyword: nchoice => GhVM|TeEn|VpCj|VpAE|CeAE|DDIE|DEMo|FeWr|HaEE|GnHu|KoAK 21:07:46 =nemelex koak 21:07:49 Defined keyword: nchoice => GhVM|TeEn|VpCj|VpAE|CeAE|DDIE|DEMo|FeWr|HaEE|GnHu|KoAK 21:07:58 yeah, it says that even if it doesn't change it 21:08:02 so someone beat me to the punch 21:14:54 !lg * t / won 21:14:55 672/25707 games for * (t): N=672/25707 (2.61%) 21:15:04 !lg * t0.22 / won 21:15:22 2148/81362 games for * (t0.22): N=2148/81362 (2.64%) 21:28:54 !tstat 5 21:28:59 !tstats 5 t0.22 21:29:01 !tstats 5 t0.21 21:29:06 !tstats 5 21:29:10 it's looking very similar to t0.22 21:29:40 it's definitely better winrate in 0.22 as I recall 21:29:50 tstats excludes bots and boring games 21:30:13 Stats after 5 days (t0.22): 1953 players, 664 runers, 349 winners, 814 wins, 24657 games, winrate 3.30%, total player time 1y+194d+9:23:32. 21:30:22 Stats after 5 days (t): 2019 players, 604 runers, 296 winners, 632 wins, 22673 games, winrate 2.79%, total player time 1y+131d+0:15:23. 21:31:03 you can play games and quit for score, and bot games are weird since the bots just die a lot 21:31:17 ah 21:31:22 so it looks like the trap changes have had a significant effect 21:31:27 !tstats 5 0.18 21:31:46 Stats after 5 days (t0.21): 1981 players, 652 runers, 343 winners, 691 wins, 26652 games, winrate 2.59%, total player time 1y+187d+13:46:12. 21:31:46 !tstats 5 t0.18 21:31:59 Stats after 5 days (0.18): 1592 players, 427 runers, 195 winners, 314 wins, 16679 games, winrate 1.88%, total player time 343d+12:59:27. 21:32:10 notable that it's still a higher winrate than 0.21 21:32:24 it wasn't at first but seems to have increased over the tournament days 21:33:05 Stats after 5 days (t0.18): 1698 players, 480 runers, 242 winners, 436 wins, 21416 games, winrate 2.04%, total player time 1y+87d+9:02:41. 21:33:36 2019 players, more than 0.22 or 0.21 so far 21:34:06 interesting that playtime and games are lower even compared to 0.21, which had lower winrate overall so far 21:34:09 but player numbers are up 21:34:41 but we keep shifting the tournament by a month or so and I'm not sure how much that has an effect 21:35:24 I think 0.19 had our all-time highest participation so far 21:35:30 !tstat 5 t0.19 21:37:05 I guess we never really figured out where the winrate shift from around 0.18-0.19 to 0.20 came from, but it's definitely lasted 21:37:12 !tstats 5 t0.16 21:37:32 yeah, and we sort of saw subsequent ones too that I don't now were fully accounted for 21:37:40 s/now/know/ 21:38:17 Stats after 5 days (t0.16): 1578 players, 641 runers, 365 winners, 675 wins, 15473 games, winrate 4.36%, total player time 1y+195d+17:30:47. 21:38:27 good old meleebug 21:38:30 t0.22 is not actually that for off 21:38:31 that 21:38:38 yeah, was just reading the meleebug logs 21:38:43 good times 21:38:58 well, one thing about that 21:39:04 is that t0.16 is partial meleebug 21:39:10 since that was fixed like halfway 21:39:16 right, but thte 5 is 5 days, right? 21:39:17 so if it were meleebug the full way, that would be higher 21:39:27 !kw meleebug 21:39:28 Keyword: meleebug => meleebug016|meleebugtrunk 21:39:29 I think it was fixed later than that? 21:39:42 !kw meleebug016 21:39:42 Keyword: meleebug016 => vlong>=0.16-b1-18-g9590bb8 vlong<0.16.0-18-g484a79e 21:39:57 %git 484a79e 21:39:57 07reaverb02 {elliptic} * 0.16.0-18-g484a79e: Don't double all player melee damage (Revert e0bdd66d8) 10(3 years, 11 months ago, 1 file, 6+ 14-) 13https://github.com/crawl/crawl/commit/484a79e65fdf 21:40:08 hrm 21:40:11 the post gives details 21:40:19 but I think it was longer than 5 days of not meleebug 21:41:06 at the end of tourney, but still winrate would have climbed higher, guess it's hard to say just how much higher though 21:41:55 !kw t0.16 21:41:56 Built-in: t0.16 => start>='2015-03-13 20:00:00' time<'2015-03-29 20:00:00' ((cv=0.16|0.16-a)) explbr= 21:42:24 Nearly all online games played during the 0.16 tournament were affected up until the fix, namely any game played from version 0.16.0 to version 0.16.0-17-g811c1d7. The fix in version 0.16.0-18-g484a79e was made on Sun, 22 Mar 14:24 UTC, however server updates took some time, and not every server was able to be updated at the same time. 21:42:57 so almost 7 days of not melee bug 21:43:28 not sure what servers updated when, I seem to recall most had pretty quickly but maybe cwz or a few others didn't 21:44:31 comment by cheuers 21:44:34 *chequers 21:44:39 This bug is responsible for 25% of my wins :| 21:46:08 heh yeah I had 5 tournament wins in 0.16, which is a lot for me 21:46:33 10% of my wins in fact...though only 3 of those register as meleebug in sequell 21:48:15 yeah, started in meleebug but transfered to not-meleebug 21:52:10 21:29:14 reaverb: and we will hopefully be migrating to aut scoring soon anyway 21:52:21 hehe 21:52:22 "soon" 21:52:31 he has a local branch for that he made maybe in 0.21 21:53:01 he stopped working to address spawns, one of the things we worked to change that version 21:53:16 there was some connection betwee the two he needed to address 21:53:39 we worked to remove spawns but he's probably not had time to work on that since 21:54:54 I guess my fixed point branch is starting to get there 21:55:59 brogue's conversion to fixed point was kind of a mess, it gave me some pause 21:56:29 though not helped by the dev kind of vanishing after the release 21:56:33 I think I started working on transporters in 0.18 but was so intimidated by travel.cc and it took me two versions of reading it off and on before I finished them in 0.20 21:57:24 my other lingering project, I also have a personal branch that half implements big monsters (would eventually subsume krakens) 21:57:36 oh, how would those work? 21:57:49 tentacle monsters are the worst 21:58:20 as in krakens/starspawns/I guess summon forest things 21:58:29 yeah, their code is somehting 21:58:32 *something 21:58:37 all the ones I implemented so far are like krakens in that they are flexibly sized 21:59:00 -!- n1k is now known as n1 22:00:31 one thing that would make me pause about undertaking a rework is that none of those monsters are very well designed gameplay wise; certainly big on flavor, but I think "pulls you closer" and "constricts" you doesn't require anything fancy 22:01:19 krakens are very cheeseable and starspawns are not particularly interesting, or at least don't need much more than a pull effect of some kind 22:02:16 mumra had a kind of crazy idea of a monster that was sort of variably sized 22:02:23 but it was a swarm where you could move into it 22:02:43 yet each movement was a chance to hit it; you would either strike the swarm or move 22:03:39 and I suppose it would surround you the entire time unless you manage to move to the edge after enough tries 22:04:02 not sure that it would work too well in practice other than being a flavor gimmick 22:05:25 one thing I had planned on trying was something a bit like that for slimes, where the bigger sizes could surround you 22:05:54 the code I wrote could also be used for actual packs 22:06:34 I'm not sure if I'll ever finish the branch, the technical challenge was interesting but it's not something crawl was ever designed to have in any sensible way 22:07:12 the tentacle implementation is up there with the lua procedural layout code in terms of mad science 22:08:20 heh, yeah I've read through it a few times and it is just kind of a nightmare 22:08:32 also a frequent source of bugs over the years 22:26:33 Hi 22:26:44 So .23 is like 20 times harder now right? 22:34:38 no 22:34:53 it's a bit harder than 0.22 though 22:35:31 I'm not even sure what you're doing that food is an issue suddenly; it's still a lot more common than it used to be 22:38:45 I am going crazy 22:39:03