00:15:04 Unstable branch on cbro.berotato.org updated to: 0.30-a0-458-g86641cc369 (34) 00:54:32 Monster database of master branch on crawl.develz.org updated to: 0.30-a0-458-g86641cc369 01:32:41 Fork (bcrawl) on crawl.kelbi.org updated to: 0.23-a0-4850-gd9e8576752 01:55:28 Fork (bcadrencrawl) on crawl.kelbi.org updated to: 0.03-1461-g08e91b2d95 04:24:18 Experimental (bcrawl) branch on underhound.eu updated to: 0.23-a0-4850-gd9e8576752 05:06:44 Unstable branch on crawl.akrasiac.org updated to: 0.30-a0-458-g86641cc (34) 11:20:30 Has anyone tried using these new AI tools that are in the news to generate new vaults or dungeon layouts? 11:20:36 (or randarts) 11:41:31 <06a​dvil> I've played a bit with what chatgpt knows about dcss, and while I'd say it knows a lot about rpg-ish games in general, it doesn't have a lot of very specific knowledge. No idea if it would have the data to generate vaults though (which is maybe a separate question) 11:42:40 <06a​dvil> example, I asked, "how do attacks of opporunity work in dcss?", and it said: In DCSS, attacks of opportunity are a type of reaction that a player can trigger when an enemy enters or leaves a square adjacent to them. This reaction allows the player to make a free attack against the enemy, without using up their action for the turn. Attacks of opportunity can be a useful way to deal additional damage to an enemy or to prevent them from 11:42:41 escaping. However, they can also be risky, as they leave the player vulnerable to counterattacks from other enemies. To use an attack of opportunity, the player must have a weapon equipped and be in a position to make a melee attack against the enemy. 11:42:52 <06a​dvil> (sorry, that'll look terrible in irc) 11:43:58 <06a​dvil> someone (with some heavy prompt engineering) got it to generate a great changelog though: https://discord.com/channels/205277826788622337/205316046230388737/1049468487657070652 11:44:12 <06a​dvil> that might have been gpt-3 and not chatgpt 11:46:59 <06a​dvil> I guess that probably needs roguelikes discord membersship 11:48:50 <06a​dvil> https://gist.github.com/rawlins/5e6c323b2d96a4cfa2d181fb1cbe3155 11:49:23 <06a​dvil> it's all still a bit uncanny valley relative to a real changelog 12:01:41 nice 12:01:50 Looking forward to praying among the stars 12:25:44 <06a​dvil> sounds plausibly like a new lugonu thing 12:25:54 <06a​dvil> wasn't there supposed to be more incentive for converting in abyss 12:37:59 03advil02 07* 0.30-a0-459-g9104f0fc43: fix: show misc evokables in quiver menu 10(73 seconds ago, 1 file, 1+ 0-) 13https://github.com/crawl/crawl/commit/9104f0fc43fe 12:51:14 Unstable branch on crawl.kelbi.org updated to: 0.30-a0-459-g9104f0fc43 (34) 14:54:10 New branch created: pull/2881 (1 commit) 13https://github.com/crawl/crawl/pull/2881 14:54:11 03Nikolai Lavsky02 07https://github.com/crawl/crawl/pull/2881 * 0.30-a0-460-gff112337aa: fix: show perma uselessness in shop and shopping list menus 10(11 minutes ago, 1 file, 2+ 2-) 13https://github.com/crawl/crawl/commit/ff112337aab3 15:02:34 <08n​lavsky> ^ it's not a new bug, but now it's more apparent, thanks to greyed out scrolls of identify 15:07:14 <06a​dvil> Ah good catch 15:08:01 <08n​lavsky> It feels like ?identify should either have a secondary function, which makes them perma useful, or there should be a daction that puffs them out of existence when everything's identified 15:08:08 <08n​lavsky> The current state when the player should keep track of the ?id's colour to know when they can throw the scroll away and get a free inventory slot doesn't seem ideal. 15:16:55 Maybe emit a message so you can set a force-more for it in your RC file? 15:16:56 <08n​lavsky> ("doesn't seem ideal" -- to put it another way, it's something that could be done automatically, w/o bothering the player) 15:17:03 "You have identified all dungeon items." 15:17:07 <05k​ate> i don't think they're significantly different to other items that can become useless (there's a bunch of things that become useless depending on the god you pick for example), and you do already get a message that they're now useless 15:17:46 Unlike useless-due-to-gods, ID can never become useful again 15:18:06 And even e.g. potion of haste is useful for Chei if you're willing to risk wrath to get the speed boost, right? 15:18:24 I'm confusing dangerous with useless, sorry. 15:18:45 <05k​ate> shields when you sac hand under ru would be comparable, we don't (and shouldn't, imo) make those all disappear from the game in that case 15:19:24 Unless you worship Beogh, then you could give the shield to a follower maybe? lol 15:20:19 <05k​ate> ?id does have the distinction that it's just a thing that comes up every game - i wouldn't be opposed to making them disappear once useless necessarily, crawl does just try to stick to the rule of making dungeon generation not depend on player state. so they'd have to still be generated as usual (and then poofed out of existence) rather than being replaced with other scrolls 15:23:13 <08n​lavsky> Is there a final message for identifying all scrolls and potions? A default force_more for it might be useful 15:23:56 <05k​ate> ah, i suppose there isn't a specific one for when you've identified all potions and all scrolls yeah, just separate messages for each 15:26:21 <08n​lavsky> yeah, items.cc has only mprf("You have identified the last %s.", class_name.c_str());, which is for a class of items 15:28:01 <05k​ate> ah and technically ?id becomes useless before you see that message i think? because the identification of the final item happens when you actually find one in-game 15:28:45 <05k​ate> so eg if you've identified every scroll, and every potion except !experience, ?id is now useless (and is correctly marked as such) but you don't get the message about identifying the last potion until you actually find an !xp 15:28:47 I thought you got identification of the final item as soon as it was the only unidentified one left 15:29:12 (since you could get the last one from the \ screen) 15:36:24 <08n​lavsky> An alternative to puffing all ?id scrolls in the inventory/shops/dungeon might be something similar to Dj's trying to pick up a book. That is, burn all ?id in the inventory and then burn any id scroll the player tries to pick up 15:37:30 <08n​lavsky> (which will inevitable bring up bug reports too 🙂 ) 15:39:40 <05k​ate> that seems like a reasonable approach yeah. might just need a message for when it happens to clarify why (maybe scrolls of identify lose their magic and disintegrate when picked up by someone who already has all the knowledge they could impart or something) 16:06:28 * Pinkbeast says "or maybe some kind of scheme where wands..." and runs away 16:23:44 Unstable branch on underhound.eu updated to: 0.30-a0-459-g9104f0fc43 (34) 16:40:23 Allow alt-firing useless ID scrolls to reveal any single tile on the map. 16:40:48 Require ID scrolls to x-v a monster. 17:01:24 <10P​leasingFungus> wow 18:55:15 <08n​icolae> just out of curiosity, do the various adjectives for unIDd artefacts in randname.txt have any correlation to tiles, like how pan lord descriptions and tiles are coded to go together. i was thinking of adding some descriptors. 18:55:37 <08n​icolae> also i have another update to randbook.txt i was planning on deploying At Some Point 19:35:00 <10P​leasingFungus> no 19:35:23 <10P​leasingFungus> there’s generally only one tile per weapon/armour type for artefacts 19:35:32 <10P​leasingFungus> randarts, that is 19:35:42 <10P​leasingFungus> doesn’t vary by description