00:33:10 Fork (bcrawl) on crawl.kelbi.org updated to: 0.23-a0-4635-gd11b0deb9c 03:23:02 Experimental (bcrawl) branch on underhound.eu updated to: 0.23-a0-4635-gd11b0deb9c 05:28:26 Unstable branch on crawl.akrasiac.org updated to: 0.29-a0-369-gd956c00 (34) 06:47:06 suggestion: you should get a free turn the first time you descend any stair, not just the first stair on a level 06:48:32 the free turn is spoilery and the optimal resulting strategy - keep fighting from the first stair even if it seems slightly bad, because you might get instakilled on the next stair (as has happened to me multiple times) - is quite unnatural 06:49:22 stairdancing is good enough as it is ;-) 06:51:20 this doesn't enable dancing, it's only the first time 06:51:20 usually there won't be anything nearby so it won't do anything useful 06:52:36 (that is, usually there won't be any monsters nearby on the new level the first time you descend a stair) 06:59:45 <10PleasingFungus> hm, i agree that the current behavior could be explained better, but disagree that it’s unnatural or bad 07:00:10 <10PleasingFungus> i think it’s great that you need to think twice about mashing < when you descend into a bad scene 07:00:53 <10PleasingFungus> “is this worse than i’m likely to see elsewhere?” is one of the bigger tactical decisions players get 07:04:15 that is nice, but I guess those are two separate things - thinking twice about < and "is this worse than elsewhere?" 07:04:15 with a free turn everywhere they'll still have the second, but not the first (as much) 07:04:46 it is nice and tactical to have to think about the first but I'm guessing most players don't think about the first because they don't know about the free turn (is it even documented anywhere, anyway?) 07:05:59 <09gammafunk> well, in terms of player knowledge, your proposal doesn't actually solve that issue 07:06:02 (and if players aren't thinking about the first, then there's no cost to losing it, and the benefit of losing it is less annoying instaloss scenarios) 07:06:09 <09gammafunk> players still have to know about the free turn with new stairs 07:06:24 <09gammafunk> what's more you have additional tracking that's harder to automate for the player 07:06:34 <09gammafunk> because it's very easy to check if you've visited a level before 07:06:41 i don't think so, it just works for free - if they go down a new stair, they might see dangerous stuff and hit < immediately 07:06:42 <09gammafunk> it's somewhat more complicated to check if a stair is new 07:07:19 ("i don't think so" re: players having to know about the free turn) 07:07:51 <09gammafunk> need more than "I don't think so", need a reason why; it's definitely just as much of a spoiler as the current system 07:08:27 <09gammafunk> as for the decision space, I think the "is this worse than elsewhere" decision is just significantly less deep without that penalty 07:10:45 I do think this is something new players get tripped up by, but the solution for that is probably clearer ui 07:10:54 i agree it's significantly less deep without that penalty, but I just don't think most players take that penalty into account today - they already live in the less-deep world 07:11:22 <09gammafunk> right, but most players don't win much, they start winning more when they do consider these things 07:11:24 hard to say, I'd expect that any player who can get a rune knows about this and at least some of the time factors it in 07:11:36 which numerically might not be "most players" ofc 07:11:54 <09gammafunk> yeah, it's fairly important info to be aware of, and certainly common knowledge among "experienced players" 07:12:30 how would a player know about this? is it documented anywhere? the way I learned about it was from the crawl community 07:13:01 i think it's reasonable to guess that the majority of players that exist aren't reading forums/reddit/discord and so could never even learn about it in theory 07:13:06 ??stairs 07:13:07 stairs[1/10]: Taking the stairs has two parts. The 'leaving' is implemented as a 'delay' - it can be interrupted. It lasts 1.0 turns. The 'entering' lasts 0.75 turns (if you've never been here before) or 1.5 turns (if you have) and is simply waited out on the other side. Haste/slow will change these durations. 07:13:39 (i for one only learned about it long after getting 15 runes for the first time) 07:13:43 I mean, I think you're right that it's not clarified in game aside from the time indicator itself, and that's something that should be fixed, but it's also not exactly hard knowledge to get a hold of 07:13:51 <09gammafunk> I can't defend the current system to the grave or anything, since just how much tatcical depth is necessary is somewhat hard to pin down without a lot of play-testing, but this aspect of stairs has been play-tested for a long, long time. I'd also lean towards better documentation/UI 07:14:19 <09gammafunk> possible pre-dcss 07:14:48 fwiw I think knowing about the extra turn when entering a new level is extremely low decision-making value 07:15:06 <09gammafunk> it's not something you think about much? 07:15:17 definitely not 07:17:03 <05kate> technically you can learn about it by seeing that the time taken display changes on the HUD when you enter a floor for the second time, yeah, certainly it's a very hidden feature though 07:17:03 like I'm certainly aware of it, but there's an extremely thin line where it matters at all 07:17:03 <09gammafunk> I've lost at least one game due to dangerous monster adjacent to second stair taken on level 07:17:03 because if a situation after going downstairs is dangerous, you should almost certainly go up instead of worrying about this 07:17:12 I'm not saying you can't lose games to bad second stairs and this turn being relevant 07:17:27 I think for new players the issue that seems to trip them up is very simple, that subsequent entries are different than the first, so they will descend into a crowd without realizing the penalty 07:17:29 I'm saying that you'll lose many more games if you refuse to use a second stair 07:17:35 <09gammafunk> elliptic: one case that stands out for me is sonja waiting near second stair on a lair level. Took second stair and banished before I could act. If I'd fought things out more on the first stair, or just escaped from first stair, decent chance I'd not have died 07:17:37 <09gammafunk> ah 07:18:02 <09gammafunk> yeah, I would never refuse, I guess, but I would think pretty hard about it (esp after that sonja experience) 07:18:07 advil: right, that's to make stair dancing worse 07:18:32 but I think making the first time you take any staircase not have this penalty wouldn't make decisions particularly more interesting or less spoily 07:18:50 yes I've had multiple experiences like gammafunk 07:19:06 (okay two) 07:19:20 instant deaths on second stair 07:19:40 hence why I think you should get a free turn on the second stair just to stop that :) 07:20:00 <09gammafunk> just made me think quite a bit harder about taking second stair and really weigh whether the situation was not salvageable in terms of getting to safety on the destination, away from first stair taken 07:20:29 just for the sake of argument, if the free turn is minor and unimportant, then a free turn on all new stairs won't hurt and will stop these annoying nothing-you-can-do deaths 07:20:39 again though, knowing about this mechanic probably shouldn't change your decision there 07:21:02 <09gammafunk> because that one turn isn't really enough of a guarantee? 07:21:11 no, because staying in a bad situation is bad 07:21:36 <05kate> i think the most i take it into account is by not dipping into downstairs during regular exploration, but i feel like removing the free turn entirely would probably be fine 07:21:39 I once explored from the first stair but was blocked by a pack of gnolls in getting to the rest of the floor, I went upstairs and back down the second stair, and then was instakilled by the gnolls because the second stair landed me in the middle of them 07:21:48 <09gammafunk> well I wouldn't stay in a bad situation, but I'd not get out of the situation by using < is what I'm saying 07:21:52 I'm just saying that with the way the mechanic currently works, knowing about it isn't particularly useful 07:23:34 (i'd also be fine with just removing the mechanic, since that would make it more clear that going down a stair is dangerous) 07:23:34 (by player experience of encountering more harder-to-deal-with dangerous situations when going down a stair) 07:24:12 also, for consideration, see advice like http://crawl.chaosforge.org/Escaping_from_(and_avoiding)_trouble#Triple_visitation which is possibly just straightforwardly bad advice given the free turn 07:24:29 I'm not sure 07:25:39 like knowing where stairs are is potentially very useful when there are dangerous awake enemies you are trying to dodge 07:25:58 I don't usually do this but it's definitely not straightforwardly bad in any sense 07:27:30 like when I am forced to retreat up my first stairs because of sigmund or something, I will sometimes make sure I know where both of the other upstairs on the level are before continuing to explore 07:27:59 <10PleasingFungus> the wiki, god bless, is full of lots of bad advice 🙂 07:28:04 that's slightly different from that advice, but it's still definitely not the case that going down unnecessary stairs is always bad or anything 07:28:08 <10PleasingFungus> don't think there's anything special about stairs in that regard 07:28:14 sure, but if the first stair is clear then it's better to explore from the first stair then go down other stairs 07:28:26 maybe? 07:30:02 like this is far from clear to me 07:30:02 one thing to keep in mind is that going down unknown stairs (if you do it at all) is safer before you've done lots of noisy stuff on the level 07:30:20 <10PleasingFungus> fwiw, speaking as a player who's worse than elliptic, i'm not sure that "if a situation after going downstairs is dangerous, you should almost certainly go up". It's sometimes right to use a consumable / god power instead 07:30:27 <10PleasingFungus> but yeah, noise matters here 07:30:42 I think stairs feel unusually safe to take since most stairs players take are the first stair to a floor, which get a free turn, which makes them safer; but usually stairs are dangerous - they jsut seem artificially safe due to this free turn 07:30:59 I mean, they are still quite safe in a lot of ways 07:31:59 <09gammafunk> I think elliptic's point was more that your decision to take a second stair or not isn't much influenced by the existence of a free turn or not on the second stair 07:32:11 yeah 07:32:12 <09gammafunk> and that the decision space is actually very similar 07:32:34 the chance of something going really wrong (like sonja banishing you) is tiny 07:33:18 yeah, that makes sense, but I'm saying your decision is probably influenced by your experience of stairs as generally safe (so, I guess I'm saying, players will play *as if* there's a free turn on the second stair) 07:33:24 my decision on whether to go upstairs is much more based on "what if I am forced to go upstairs on the other two stairs too and then really don't want to go back down the first one since things are worse the second time" 07:33:33 (especially if they don't know about the free turn mechanic, which I think most players do not) 07:33:42 rather than any consideration of actually dying to new stairs 07:34:01 again, I think my decision would almost never change 07:34:37 that free turn has to be significant in <1% of times 07:36:32 <12ebering> I think the most significant impact of the free turn is on the design of certain ambush vaults but even then it's not that big 07:36:32 to be clear I'm not necessarily defending the current status quo, I'm just saying that I don't think knowledge of the mechanic is particularly useful or interesting from a decision-making perspective 07:36:32 <12ebering> (would be sad to lose the aesthetic experience of V:5) 07:36:54 don't think it's relevant for V:5 07:37:10 <09gammafunk> maybe ebering means the perfect ring of vault guards? 07:37:11 <10PleasingFungus> it means that all the vault guards will be standing in formation when you enter v:5 07:37:13 <09gammafunk> all lined up, so pretty 07:37:17 oh, aesthetically 07:37:25 yeah, that's fair 07:37:32 <09gammafunk> it is because he is a tiles player through and through 07:38:11 I do kinda think the <1% times that it does matter are extremely memorable (in a bad way) events 07:38:43 which is exactly what prompted this discussion I suspect 07:38:50 <10PleasingFungus> re removing the free turn, i agree it'd rarely have much effect, but the effect it would have would mainly be to create more 'feel bad moments' - bad things happening to the player without any chance to react. certainly rare, but it's not clear to me what the upside would be 07:38:55 yes 07:38:57 <10PleasingFungus> wow, i'm too slow and advil beat me 07:39:05 <10PleasingFungus> if only i'd had a free turn to respond 07:39:08 lol 07:39:09 well, you expressed it better 07:39:22 <09gammafunk> well you took the second stairs into the dev channel, via discord, so your fault 07:39:35 advil: yes, that's what prompted it :) so that's why I started by "let's add free turns to all new stairs" 07:39:39 irc superiority proven 07:40:18 <10PleasingFungus> lol 07:40:29 (second stair surrounded by grum and his pack, first turn takes me to half health and lethally poisons me) 07:40:41 so the issue with adding free turns to new stairs is that we don't currently record that information for the player 07:40:42 <10PleasingFungus> aw, that's not so bad 07:40:49 and it wouldn't be useful for anything else 07:40:54 <10PleasingFungus> that means you still have half your health left! 07:41:08 <09gammafunk> elliptic: I think we do, right, in that we record in the UI whether stairs are new? 07:41:09 unless you would fail to get the new turn if you've seen the other side of the stairs but not used them, which seems... weird 07:41:13 <10PleasingFungus> gammafunk: no 07:41:15 <10PleasingFungus> well 07:41:22 we record whether or not you have seen the other side of stairs 07:41:26 not whether they have been used 07:41:27 elliptic: I think that would be fine actually! 07:41:28 <09gammafunk> ah, true 07:42:56 <10PleasingFungus> it's not clear to me what a "new stair" means 07:42:56 <09gammafunk> heh, in that case, free turn removed! 07:42:56 <09gammafunk> it's really awkward yeah, because shafts 07:42:56 elliptic: that would serve the purpose of preventing instadeath on descending an unknown stair 07:42:56 <09gammafunk> whole host of bad issues probably 07:42:56 I forget whether forgetting the map messes with that 07:42:56 <09gammafunk> wow, is docs a new commit prefix 07:42:56 haha it's a standard one I think 07:42:56 it's just a bit awkward because that stair coloration is intended to be an interface thing 07:43:03 <10PleasingFungus> docs! is what you say when you're hurtin real bad 07:43:08 not a (minorly) gameplay-relevant thing 07:43:08 so you could actually unify the mechanism to "entering a stair where you haven't seen the other end, gives you a free turn" :) 07:43:49 I think that might be a bit more confusing if anything 07:44:11 though one could just directly indicate in text on a stairs whether you would get first move 07:44:14 <09gammafunk> advil: I guess I might replace the word "move" with "act" there, although I doubt people are confused by that 07:44:21 <09gammafunk> in that commit, I mean 07:44:27 sure, I wrote that a bit quickly 07:44:37 ew, I definitely don't like making this (extremely minor) mechanic explicit like that 07:45:21 since again, I don't think it is very useful to know about it - it's the sort of thing where having it invisibly prevent instadeaths if possible is nice, but people don't need to know it is doing that 07:45:34 <10PleasingFungus> is it harmful to mention? 07:45:39 I think so? 07:45:55 interface clutter of limited decision value is generally bad, IMO 07:46:31 fwiw I just commited a sentence to the manual about this, I guess maybe you're talking about the ui suggestion though 07:47:00 yeah, I guess maybe if you meant the description when examining the stairs as a feature on the ground that's fine 07:47:05 I think new players *really* do need to know about the stairdancing penalty 07:47:16 I don't 07:47:17 Unstable branch on crawl.kelbi.org updated to: 0.29-a0-370-gb5314be063 (34) 07:47:21 think it's very easy to learn in practice 07:47:25 like many things 07:47:25 and you can't describe that accurately without describing this mechanic 07:47:34 well, I think we disagree then 07:48:15 03advil02 07* 0.29-a0-370-gb5314be: docs: document stair delays (catern) 10(13 minutes ago, 2 files, 8+ 3-) 13https://github.com/crawl/crawl/commit/b5314be063d2 07:49:29 like if a player is new to dcss, I'm not going to tell them "don't stairdance, monsters will get free hits on you" 07:49:29 I'm going to let them try that themselves and find out how it works (and sometimes stairdancing is still very good) 07:49:29 conveniently, that's not what I think they need to know 07:49:29 seems bizarre to me to say that new players "*really* do need to know", they clearly do fine without being told 07:49:29 <10PleasingFungus> do they? 07:49:29 <10PleasingFungus> they seem to die an awful lot 07:49:29 what do you mean by your statement then? 07:49:29 agreed 07:49:32 <10PleasingFungus> (i agree that we shouldn't clutter up the ui with more markers on stairs or w/e) 07:49:55 new players need to know that if they stairdance, monsters get a free move and they should expect more damage 07:49:58 <10PleasingFungus> (i agree this is a minor mechanic & so should only show up in low-impact contexts like the manual & a hints mode message) 07:50:06 I still disagree very strongly 07:50:11 ok 07:50:12 <10PleasingFungus> (but I do think we should document most mechanics the player can interact with) 07:50:12 I think putting this in the manual is great 07:50:25 but that's extremely different from "new players need to know..." 07:51:15 <09gammafunk> ok how about this, intermediate players really need to know it 07:51:20 if I was making a list of 100 things to tell a new player, this wouldn't come close to making it 07:51:50 <09gammafunk> you don't have a list of 100 things to tell a new player, this is all conjecture!!! 07:52:27 you caught me 07:54:07 like yes, I would hope that a player learns at some point that if they go <><> with monsters around then they will take damage while doing so 07:54:26 just as one thought, if this is actually irrelevant, then some players (me) will start playing suboptimally after learning about it, and try harder than they should to explore from the first stair instead of going to a second stair 07:54:28 but I think this is extremely easy to learn in practice 07:55:54 hm, the uv4 guide doesn't mention it and the patashu guide does (each of these probably has more than 100 things by far in them) 07:55:54 <10PleasingFungus> think learning how to evaluate the impact of different game mechanics is part of the natural process of becoming better at the game 07:55:54 <10PleasingFungus> as long as we don't phrase things too dramatically, i think it's fine 07:55:54 <10PleasingFungus> (and i don't think advil's additions were particularly dramatic?) 07:55:54 <10PleasingFungus> it's not like he added a warning for "are you sure you want to go up? you won't get a free turn when you come back down again!" 07:55:54 <10PleasingFungus> which would be a great custom rc mechanic for goofballs, btw 07:55:54 "type "yes" to go up" 07:56:02 <10PleasingFungus> exactly 🙂 07:56:40 I think the addition to the manual was great as I said, but I also think viewing the manual as "things new players need to know" is not how I see it at all 07:57:11 I see it as "if a player wants to read a little about how x topic works, they can do so there" 07:57:19 <10PleasingFungus> yeah, i feel like the disagreement between you and advil might be more about how we discuss it / how we define "need to know"? 07:57:25 <10PleasingFungus> how we discuss it here, that is 07:57:26 (should it be in the tutorial?) 07:57:38 <10PleasingFungus> seems hard to tutorialize 07:57:45 <10PleasingFungus> and not important enough 07:57:47 I considered it but the tutorial doesn't need more stuff probably 07:57:49 <10PleasingFungus> the tutorial is fairly short right now 07:57:50 <10PleasingFungus> yeah 07:57:55 also harder to modify tbh 07:58:26 so if the tutorial is our "100 things a new player needs to know" I guess I did agree with elliptic 07:58:53 <10PleasingFungus> advil can correct me if i'm misrepresenting him, but i don't think advil was saying that the stair-climbing thing is something that new players need to learn ASAP. rather, I think his point is that it's something they don't start out knowing & that isn't really presented otherwise, so to truly master the game, they "need to learn it" at some point 07:59:12 yes, I think that's a good summary 07:59:22 sure, and I'm saying that they can learn it by doing <> a couple times 07:59:25 <10PleasingFungus> "new player" here being used as "someone who's on the path to learning about the game", rather than as "someone who just launched the game for the first time" 07:59:49 <10PleasingFungus> well, i mean, that shows that later descents take time, but not that the first one doesn't? you'd have to really be watching for it to catch that 07:59:54 (learn the "monsters can hit you if you stairdance" part, I mean) 08:00:02 <10PleasingFungus> and empirically, it does seem like many learning players miss these things! 08:00:07 again, I don't think knowing that thing is important at all 08:00:10 <10PleasingFungus> of course, many learning players are going way too fast and not reading 08:00:18 <10PleasingFungus> so hints mode & manual won't help there 🙂 08:00:20 knowing what happens when you <><> is important 08:00:22 tbh I think "new players" in this sense usually learn this one the hard way by dying, I know I did back in the day 08:00:25 and easily learnable 08:00:25 am i out of touch? no, it's the players that are wrong! :P 08:00:40 right, dying is a great way to learn 08:00:45 <10PleasingFungus> yep! 08:00:55 <10PleasingFungus> that’s what rogues are all about 08:02:23 it's like learning that sigmund is dangerous 08:02:23 <09gammafunk> hey I bet hints mode mentions sigmund 08:02:23 adding "type yes" prompts for various trivial dangerous things would be a good april fool's fork :) 08:02:23 "do you really want to move towards sigmund? type yes" 08:02:23 <10PleasingFungus> think that’s funny for about 3 seconds 08:02:30 <10PleasingFungus> better in irc than implementation 🙂 08:03:26 anyway, returning to the topic of giving the free turn whenever you use a staircase with unknown destination 08:03:54 I think that would probably be fine and a minor improvement if it isn't a major code burden 08:04:13 one thing that's a little weird is how it interacts with shafts/hatches 08:04:24 currently you get a free move for them too if it's to a new level 08:04:53 but I don't know how that would be "unified" exactly with the staircase rule 08:06:52 <10PleasingFungus> ty advil for making the docs changes, btw! swift and decisive 08:08:49 <10PleasingFungus> i have no strong opinion on the unknown stair turn change 08:08:49 <10PleasingFungus> i guess you could consider shafts and hatches to always have an unknown destination? 08:08:49 <10PleasingFungus> that seems reasonably consistent 08:08:49 shafts at least 08:08:50 hatches have a fixed destination, right? 08:08:54 yeah 08:09:06 <12ebering> iirc it might be an awful code burden, the place where the time taken is set is in load level which might not know enough about how you got there? 08:09:07 though you don't know it until you enter it 08:09:43 <10PleasingFungus> could always set up some temporary global state 🙂 08:09:53 <10PleasingFungus> a lil unwind var, as a treat 08:10:29 it does have the specific stair feature taken 08:10:38 <10PleasingFungus> yeah 08:10:51 but it might still be slightly tricky to use that info 08:11:30 also I can't swear that there are no horrible hacks with that param, would have to check 08:11:44 <10PleasingFungus> there definitely are some weird hacks there 08:11:57 <10PleasingFungus> check out eg how ignis rising flame works 08:12:02 <10PleasingFungus> uh, sorry 🙂 08:12:09 lol 08:12:55 <10PleasingFungus> code burden aside, i guess i feel this change would slightly reduce the impact a specific tactical decision (“go up or fight here”), but would reduce feelbads, make things align better with what seems to be common player intuition, and make the game very slightly easier 08:13:09 upside of this potential change is that even if it doesn't matter it would present as increasing fairness, especially around shafts 08:13:12 yes 08:13:18 that time you beat me 08:13:21 <10PleasingFungus> so that seems net positive to me, assuming the code isn’t a nightmare 08:13:24 <10PleasingFungus> high 5 08:15:15 is there currently the stairdancing penalty if you get shafted into an already seen level? 08:15:15 <10PleasingFungus> great question 08:15:15 I would assume so? 08:15:15 looks like yes 08:15:15 <10PleasingFungus> i don’t even remember where the penalty lives in code 08:15:15 <10PleasingFungus> i remember failing to find it in the past 08:15:15 which I think is fine (it's rare), the weird thing with shafts is what happens when you skip a level and go upstairs into the skipped level 08:15:15 !source files.cc:2191 08:15:15 https://github.com/crawl/crawl/blob/master/crawl-ref/source/files.cc#L2191 08:15:15 that's where the penalty is removed 08:15:15 I forget whether you get the penalty then 08:15:15 <10PleasingFungus> right, ty! 08:15:32 <10PleasingFungus> i believe you don’t currently get penalized in that case 08:15:38 <10PleasingFungus> same as delver, right? 08:16:04 which case? 08:16:12 going upstairs into skipped levels 08:16:14 <10PleasingFungus> hm, clearly the solution is for me to get to work implementing hellmo’s “ironman mode” 08:16:38 <10PleasingFungus> this discussion is much simpler there 🙂 08:16:39 I don't know whether those levels have already been created in the appropriate sense 08:17:17 ah yeah `just_created_level` should be true on any non-entered level regardless of pregen 08:17:46 that's good 08:17:57 intended to have the same behavior as before pregen 08:19:12 <10PleasingFungus> yeah, many downward levels will be pregenned too, right? 08:19:14 <10PleasingFungus> all? 08:19:21 <10PleasingFungus> since we generate in batches 08:20:08 I guess one issue with giving protection on all unknown staircases is that it makes that "try all the stairs before exploring the level" strat more tempting 08:21:37 and I'm not sure whether or not we want to encourage that? 08:21:37 <10PleasingFungus> i think that’s still a cost, right? 08:21:37 <10PleasingFungus> since now you’re giving up the free turn on later levels 08:21:37 ? 08:21:37 <10PleasingFungus> and risking getting shafted extra deep 08:21:37 later levels? 08:21:37 <10PleasingFungus> like uh 08:21:37 <10PleasingFungus> you’re on d:2 08:22:12 <10PleasingFungus> you take each stair down to d:3 and back 08:22:55 <10PleasingFungus> but even the best one still has a gnoll next to it, and when you go back to it, you don’t have a free turn anymore 08:23:06 <10PleasingFungus> does that make any sense? 08:23:30 <10PleasingFungus> not sure if we’re talking about the same tactic 08:23:39 I mean, the idea is that you do this after you fully explore D:2 08:23:51 <10PleasingFungus> yes, that’s what i’m talking about too 08:23:54 if all the stairs are bad, you would have gone and visited all of them anyway 08:24:12 <10PleasingFungus> at present? 08:24:22 like the strat doesn't mean "do >< with no fighting" 08:24:33 <10PleasingFungus> oh, i see 08:24:48 it means "instead of starting to explore D:3 from a clear stairs, go back up and visit the others" 08:25:32 I think it's probably fine, the strat is a bit tedious and probably optimal if you don't have piety decay, but not a big deal 08:25:49 already have tedious optimal things like not autoexploring, etc 08:28:08 there's also still some downside (more likely to have sigmund wandering around the level) so it isn't 100% clearly optimal 08:28:22 <10PleasingFungus> gotcha 08:28:23 <10PleasingFungus> agree 08:29:51 (also, iirc even with the free turn it's possible for monsters to hit you as you go back up the stairs? just a lot safer) 08:35:15 <10PleasingFungus> yeah i think so 08:36:54 <09gammafunk> yeah they can hit you if e.g. adjacent or if not adjacent and sufficiently fast 08:40:55 <10PleasingFungus> saw some complaints about getting parad on stairs while heading up, or harpoon shotted 08:40:55 <10PleasingFungus> harpooned 09:26:36 While this obviously adds tedium and wouldn't be fun, could always allow something janky like not having any free turn and instead allowing you to "peak" through a staircase where you get a chance to look around but not act while the monsters get a roll to see if they notice you for when you come down for real or something 09:27:00 (The reason I think that'd be bad is because it could encourage peeking every stair and then picking the "safe"/"best" one) 09:27:18 Also FWIW I had no idea that the free turn mechanic existed, so props to a'dvil for adding it to the docs :) 09:56:58 <06advil> I gotta say I'm now actually somewhat confused about whether what I wrote is accurate 09:58:31 <06advil> I recompiled to always use the new level time and test some things out, and even normal speed monsters that are adjacent and awake seem to get the first move 09:59:18 <06advil> maybe it plays out differently if they don't yet have a foe? 10:01:20 <06advil> that seems to be right 10:05:00 <06advil> ok yeah, I think I just needed to make sure to test it with their foe info reset rather than repeatedly going up/down 10:27:33 <06advil> heh, also in the category of "maybe this doesn't matter" but if you switch stairs you're guaranteed a call to shake_off_monster 10:27:46 <06advil> *shake_off_monster 10:28:15 <06advil> I'm not sure I knew that except at some very hazy intuitive level 10:28:31 <06advil> er, shake_of_monsters is what I keep trying to type 10:28:35 <06advil> loll 10:28:41 <06advil> shake_off_monsters 10:28:54 <10PleasingFungus> the thing i was vaguely wondering about changing is the code for wiping out enemy statuses 10:29:08 <10PleasingFungus> the massive awful switch case in timed-effects.cc 10:30:48 <10PleasingFungus> iirc it triggers if you’ve been offlevel for over 10 turns? 10:30:48 <10PleasingFungus> the threshold there feels odd 10:32:33 <06advil> looking at this I think that's because it runs the timeouts at turns / 10? 10:32:57 <06advil> if (turns >= 10 && mon.alive()) mon.timeout_enchantments(turns / 10); 10:33:10 <06advil> I don't mean to say that's not weird but it would be pointless to run before turn 10 10:37:17 <06advil> v confused about what that argument (int levels) really means 12:17:54 <10PleasingFungus> i think it might still have effects at 0 12:18:09 <10PleasingFungus> afk at present but can check later 12:48:31 looks like some ench will get deleted regardless of what that number is 12:48:43 <10PleasingFungus> yeah, a bunch of em, right? 12:49:05 <10PleasingFungus> it came up when someone was talking about boulder beetle rolling 12:49:16 <10PleasingFungus> they were surprised that they went up and down the stairs and a boulder beetle was still rolling at em 12:50:10 <10PleasingFungus> hm, i don't think there's anything very consistent going on here, actually 12:50:40 that one does seem weird but a lot of these don't seem obviously like you should be able to force a timeout with stairs? 12:50:45 <10PleasingFungus> not clear to me that there's a reason for some stuff to be in the lose_ench_levels(entry.second, levels); case and some to be in the lose_ench_levels(entry.second, levels); case 12:50:46 like berserk 12:50:54 <10PleasingFungus> what do you mean? 12:51:09 <10PleasingFungus> also, hm, i wonder how this interacts with duvessa permazerk 12:51:25 should berserk time out immediately on reentry of a level? 12:51:37 <10PleasingFungus> should it time out if you wait 10 turns and then re-enter a level? 12:52:03 <10PleasingFungus> should players know that there's a difference between waiting 10 turns and waiting 9? is there a mechanical cost involved to waiting 10 turns? 12:52:18 <10PleasingFungus> with waiting 10 turns offlevel, that is 12:53:53 I would have guessed that berserk should work like the ones that call `lose_ench_levels` 12:53:53 <10PleasingFungus> sure, but i guess i'm saying: why? 12:54:10 <10PleasingFungus> and for that matter, there's some extra weird stuff there... regardless of everything else I was saying, word of recall seems like it shouldn't need 10 turns to expire? 12:56:28 <10PleasingFungus> i think it's reasonable to argue that dcss should have a tighter clock, but it's hard for me to imagine 10 turns of offlevel resting being a meaningful cost anytime soon 12:56:30 because it lasts some time and that's what those calls seem to be handling 12:57:15 <10PleasingFungus> i guess my perspective is that, from a gameplay / "optimal play" perspective, the 10 turn threshold is "spoilerish" and doesn't add anything 12:57:38 <10PleasingFungus> it just surprises players in the rare cases that they go up and right down again without resting 13:01:43 I still think this is basically an artifact of the fact that "levels" decrease in increments of 10 turns 13:02:15 not trying to argue that isn't weird 13:02:30 fwiw this is from the famous "Initial revision" 13:03:57 so if there's coherent intent it was just for things in that category (a much smaller list back then) to take some time 13:04:33 in the commit I'm looking at that list was berserk, fatigue, tp confusion, held 13:04:42 tp, confusion 13:07:26 and tp also fires 13:07:39 slightly amused that the trick of forcing a blink at the end of confusion still exists 13:09:47 <10PleasingFungus> it’s a great trick 13:09:55 <10PleasingFungus> i approve 13:10:15 <10PleasingFungus> anyway, maybe i’ll change some of this at some point 13:52:28 what is this blink confusion thing 14:15:01 oh, when confusion duration is ended in this way 14:22:39 mhtbhm (L9 MiBe) ASSERT(defender->alive()) in 'fight.cc' at line 254 failed. (D (Sprint)) 14:24:40 yes, I take it it's just a way of simulating some confused movement over time? 14:25:06 !lm mhtbhm sprint crash -log 14:25:08 2. mhtbhm, XL9 MiBe, T:1758 (milestone): https://cbro.berotato.org/morgue/mhtbhm/crash-mhtbhm-20220501-212238.txt 15:23:27 Unstable branch on underhound.eu updated to: 0.29-a0-370-gb5314be063 (34) 16:20:07 03PleasingFungus02 07* 0.29-a0-371-g91724d5: Standardize screamitis conditions (ardl) 10(4 minutes ago, 4 files, 12+ 7-) 13https://github.com/crawl/crawl/commit/91724d5b3990 16:48:05 Unstable branch on crawl.kelbi.org updated to: 0.29-a0-371-g91724d5b39 (34) 22:34:47 Unstable branch on crawl.develz.org updated to: 0.29-a0-371-g91724d5b39 (34) 22:56:48 Windows builds of master branch on crawl.develz.org updated to: 0.29-a0-371-g91724d5b39 23:54:59 Monster database of master branch on crawl.develz.org updated to: 0.29-a0-371-g91724d5b39 23:56:29 Unstable branch on cbro.berotato.org updated to: 0.29-a0-371-g91724d5b39 (34)