00:25:55 03CrawlOdds02 07* 0.35-a0-759-g3e0666f4fb: Fix seed stability for unrand weapon generation 10(36 minutes ago, 1 file, 24+ 9-) 13https://github.com/crawl/crawl/commit/3e0666f4fb9a 00:59:03 03CrawlOdds02 07* 0.35-a0-760-g1e7925b4b1: Preserve feature flavour on feature swap 10(12 minutes ago, 1 file, 4+ 0-) 13https://github.com/crawl/crawl/commit/1e7925b4b13b 03:34:06 Experimental (bcrawl) branch on underhound.eu updated to: 0.23-a0-5261-gd9800d219b 05:30:28 <11O​dds> I've been thinking about enabling items to grant slots which are melded by the current form (gloves on the fisticloak in granite form, etc), which I think people agree is desirable but a bit tricky in the current code. There are two approaches we could take here: 1) Give these items their own slot (so there is a particular pair of gloves on the fisticloak, etc) 2) More complicated slot bookkeeping while keeping these slots the same I'd 05:30:29 be inclined to do 1) - I think it's a bit clearer, especially in talisman cases, and substantially easier to get right (we've had plenty of bugs in this area and still have at least one with melded unrands blocking real slots). It is also a little bit worse for the player - you lose flexibility if you have cursed/fragile/etc items on unrands you want to remove or meld. But very interested in other opinions here - especially I suspect @dracoomega has 05:30:29 thought about this before! 05:38:13 <04d​racoomega> I have, in fact, though about this a bit in past. (It's one of those things I've meant to get around to for a long time, but never have.) And the idea of a separate slot type (like SLOT_GLOVES_EXTERNAL for fisticloak) was also the direction I was leaning towards. I agree that the implementation of it seems the least likely to be confusing or tangled (and there's a bit of precedent with how fortress crab's two-hander slot works.) 05:38:14 Though regarding it being worse for the player about what goes on which slot, I had vaguely leaned towards the game invisibly shuffling stuff between slots so that you didn't have to worry about what was on what. (In a similar way what you don't need to worry about exactly which ring is on the macabre finger. ...that may have some issues with melding, though, if it's essentially picking an arbitrary thing to not meld. Since, in that case, you can't 05:38:14 just equip or unequip stuff until you get the 'right' item not melded. Unsure if there's a more UI-friendly way to go about that... 05:44:24 <11O​dds> Yeah, the player will always already needs to worry about what goes on which slot in the case of melding (whether that's direct melding of an item or via a chain unrand). But we could perhaps do a reshuffle in the specific case that the player is losing slots (either to melding or permanently) and has some free slots 05:45:08 <08w​ormsofcan> the game already forces you to decide what's on extra slots, right? 05:45:23 <08w​ormsofcan> like if you have 4 rings with vainglory the game doesn't let you remove it without picking 2 rings 05:46:08 <11O​dds> But it does meld it while melding whichever two rings you last equipped (even if in fact you equipped one of them on your fresher finger necklace) 05:46:09 <08w​ormsofcan> haven't checked what happens if you curse all the rings though 05:47:06 <08w​ormsofcan> although, if the melding is automatic, that gets me wondering 05:48:04 <08w​ormsofcan> are there any undesirable effects that can happen from a meld and does the game warn them? only thing I can think of is stat zero and that's less of an issue currently 05:48:33 <11O​dds> Yeah we deliberately don't warn for stat zero on any form of equipment removal 05:49:23 <11O​dds> It will stop you melding flying over lava 05:49:38 <11O​dds> (Directly or through a chain like vainglory -> ring of flight) 05:50:07 <08w​ormsofcan> is that a "you will lose flying" type warning or "you might lose flying" type warning 05:50:24 <11O​dds> It's a "you can't do this at all" type warning 05:50:35 <08w​ormsofcan> so the former 05:50:42 <11O​dds> Yeah 05:51:35 <11O​dds> (The player can see which rings will come off - it's whichever are listed last - but that's not explained) 05:55:44 <08w​ormsofcan> oh we don't give the player a warning for melds that remove ^contam 05:56:07 <11O​dds> How can a meld remove rather than meld ^contam? 05:56:08 <04d​racoomega> Melding that doesn't give you contam, though 05:56:14 <08w​ormsofcan> sphinx with a ^contam barding 05:56:58 <11O​dds> Ah, that's a removal not a meld (and we should probably warn for it, yeah) 05:58:38 -!- hatvik_ is now known as hatvik 05:58:41 <04d​racoomega> Anyway, I wonder if some of the issue of 'Form melds 1 of 2 items the player is wearing, and the unmelded one isn't the one the player wanted' could be addressed by just letting you unequip melded items? So that you could, in fact, extricate the proper piece of gear you wanted and put it on your fisticloak, even if it got swallowed up by statue form or something? 05:59:58 <11O​dds> For the main case of voluntary forms we could also just ask? 06:01:36 <04d​racoomega> The UI for that sounds... complicated, given that it's entirely possible to be like 'Pick 2 out of these 4 things and then also one out of these 3 things' 06:01:57 <04d​racoomega> Which sounds like an awkward chain of prompts for putting on a talisman 06:02:07 <08w​ormsofcan> the problem with asking is that if the player doesn't care most of the time it's more tedious than not 06:03:12 <11O​dds> The cases of melding giving you a choice will be pretty rare right? And multiple choices even more so 06:04:02 <11O​dds> This seems mostly symmetric with removing zonguldrok removing vainglory removing two rings 06:04:38 <08w​ormsofcan> well, vainglory is guaranteed to happen no more than once 06:05:19 <11O​dds> Sure.... I guess as a player I'd be much happier not having to manoeuvre rings into the right slots (I usually do care which get melded!) 06:06:24 <11O​dds> (Was surprised to discover that chain prompts me once in the message window and once in a popup) 06:06:24 <11O​dds> https://cdn.discordapp.com/attachments/747522859361894521/1531286633951793353/image.png?ex=6a68a94f&is=6a6757cf&hm=9f5187f5b6c1fadda345668b57364293a1c700989529abd783a49c0948b05485& 06:06:51 <08w​ormsofcan> yeah it's surprisingly involved 06:08:18 <08w​ormsofcan> dracoomega's suggestion of just letting the player unequip as if they weren't melded seems fine, the only problem is explaining the flavour 06:08:44 <04d​racoomega> I mean, you can... will them out of your body, I suppose? 06:09:12 <08w​ormsofcan> yeah i'm just saying flavour justification isn't a big deal 06:09:26 <04d​racoomega> A little odd to get them both at once, but it basically gives a popup if there's more than 2 (or is it 3?) things 06:09:46 <11O​dds> It sounds like overall we're leaning towards: * Under the hood we have unrand specific slots * These aren't visible to the player * Some combination of slot reshuffling, unequipping melded stuff, prompting to make this work intuitively I think the main alternative is: * Under the hood we have unrand specific slots * These are visible to the player * Tough luck if you don't like where your items are, you must unequip stuff to move it 06:12:20 <08w​ormsofcan> how does this square with ash btw 06:12:55 <08w​ormsofcan> I assume we wouldn't let the player swap meld/unmelded status of cursed items 06:13:52 <04d​racoomega> In the old equipment system, you did have very specific individual slots for stuff like this (with the macabre finger having one specific ring) and there was actually a fair bit of code burden managing just that (and things like Sac Hand ring removal). I actively opted to avoid the player having to worry about exactly what was on which finger anything was on, and am generally inclined to avoid going back that way. 06:14:37 <04d​racoomega> Ash is awkward here, alas. (On the other hand, shapeshifting with significant melds is generally somewhat bad with Ash, so maybe it's less likely to come up? >.>) 06:16:31 <11O​dds> Cool, good to know the history... I'll have a crack at the version where the player doesn't have to care what's going on. I'm still not 100% sure that prompting isn't a good fix here (it would work well for the Ash case) 06:17:04 <04d​racoomega> ....maybe you're right. It is much better there. 06:17:27 <11O​dds> I think I'll get more confident on this once I'm deeper in the code 🙂 06:18:03 <11O​dds> (One especially awkward edge case for all this is serpent form <-> zonguldrok, where we have two hats for very different reasons) 06:19:58 <04d​racoomega> It... might be, yeah. (It's funny since the desired end behavior is obvious, but the inbetween steps might be tricky) 06:21:30 <04d​racoomega> I wonder if it's possible to 'float' equipment that is slated to be removed, and then check slots after the transformation is simulated and see if there's equivalent slots to put them in 06:21:43 <08w​ormsofcan> wait does that work smoothly for ash currently? pretty funny if it does 06:22:08 <11O​dds> No there's definitely some problem with serpent and the skull at the moment 06:22:30 <11O​dds> I can't remember the sequence, but you can end up in serpent form with a melded hat and unable to equip two 06:23:08 <08w​ormsofcan> yeah it looks like one getes melded 06:23:14 <11O​dds> (You can also actually just end up in medusa form with melded vainglory + 2 rings and unable to equip any more rings) 06:23:28 <08w​ormsofcan> pro tech btw 06:23:36 <11O​dds> (Also why is it medusa form not gorgon form?) 06:23:49 <08w​ormsofcan> isn't it not an actual gorgon 06:23:57 <04d​racoomega> That surprises me, actually. Why would it meld things you have slots for? 06:24:23 <11O​dds> If you have 4 rings, meld vainglory, and remove the two remaining rings, then the melded ones block the slots 06:24:36 <04d​racoomega> Oh, okay. I thought you meant if you had only 2 on 06:24:46 <11O​dds> Ah no, that path works fine 06:24:54 <11O​dds> I think you can also swap the remaining two, so it never actually comes up 06:24:59 <04d​racoomega> (And yes, it's a medusa in the sense of 'jellyfish'. Well, it's a double-meaning, but it is a double-meaning.) 06:25:22 <04d​racoomega> https://en.wikipedia.org/wiki/Medusozoa 06:25:32 <11O​dds> Oh! Huh! 06:25:37 <04d​racoomega> a medusa typically being an umbrella-shaped body with stinging tentacles around the edge 06:25:57 <11O​dds> Right, if I'd read the talisman description all this would have been clear 06:26:36 <11O​dds> Yeah, something like this should be doable 06:30:14 <04d​racoomega> That bug, at least, seems like it might be simpler to solve? Just don't count melded items against the slot limit if the slot itself is not melded overall? 06:30:36 <04d​racoomega> Possibly some gotcha involving merfolk tail stuff, though. (That one's ability to 'stack' with other transformations has been repeatedly tricky) 06:30:43 <11O​dds> Yep, I think it would be fine to solve independently 06:30:58 <04d​racoomega> Don't want to let the player accidentally put on more boots every time they grow a tail 😛 06:31:18 <11O​dds> Yeah if boots are ever on either end of a slot giving unrand I predict disaster 🙂 09:07:22 -!- hatvik_ is now known as hatvik 09:29:03 -!- hatvik_ is now known as hatvik 15:16:19 03Medrano8302 07* 0.35-a0-761-g8bf6446ad2: Update Android SDK and Gradle 10(31 minutes ago, 6 files, 19+ 6-) 13https://github.com/crawl/crawl/commit/8bf6446ad21b 15:43:44 Unstable branch on underhound.eu updated to: 0.35-a0-760-g1e7925b4b1 (34) 17:10:51 New branch created: pull/5351 (1 commit) 13https://github.com/crawl/crawl/pull/5351 17:10:51 03Aliscans02 07https://github.com/crawl/crawl/pull/5351 * 0.35-a0-762-gfc2f18dfc1: Hide tile-only keys in the level map help outside tile mode. 10(42 minutes ago, 1 file, 4+ 2-) 13https://github.com/crawl/crawl/commit/fc2f18dfc171 19:22:13 <09h​ellmonk> well I got fuzzed throwable durability working (for a given value of working), but I am a bit worried about internal state tracking again 19:22:21 <09h​ellmonk> specifically for stuff like gl's starting javelin 19:24:00 <09h​ellmonk> wondering if we should do some kind of falling apart/worn/slightly worn descriptors like we used to have for nets 21:01:08 <04d​racoomega> Hmm... that sort of 'percentage intact' thing feels a bit superfluous/noisy to me when it comes to having a bunch of throwables (since the insurance that you're not going to immediately run out is in having more of them afterward). Are you worried about people counting how many times they've hit something with it, to get an estimate of how many throws might be left? 21:04:55 <04d​racoomega> (Though incidentally, the single javelin you start with has always felt quite sad to me, since it isn't even very good with low throwing skill) 21:05:07 <04d​racoomega> Some of that is it having a decent chance to just break immediately, of course 21:05:11 <04d​racoomega> (Currently) 21:07:01 <09h​ellmonk> I am worried about counting throws in earlygame on single javelin / single curare etc 21:07:21 <09h​ellmonk> seems quite relevant to know that my one curare dart could mulch vs not 21:08:05 <09h​ellmonk> lategame I don't think it will matter, I removed the mulch bonus damage and penetration stopping 21:09:51 <04d​racoomega> Maybe it's fine to display this if you only have one left?? 21:09:57 <04d​racoomega> And not bother otherwise? 21:10:19 <09h​ellmonk> maybe, idk 21:10:28 <09h​ellmonk> or like at some other low stack threshold 21:14:12 <09h​ellmonk> oh, could make an rcfile setting and only display at 1 by default maybe 23:12:35 Unstable branch on cbro.berotato.org updated to: 0.35-a0-761-g8bf6446ad2 (34)