00:10:58 Stable (0.33) branch on cbro.berotato.org updated to: 0.33-b1-1-ga045eee6d5 03:33:49 Experimental (bcrawl) branch on underhound.eu updated to: 0.23-a0-5261-gd9800d219b 09:11:26 <09gammafunk> So Zigs don't have enough ghosts and you want to put some of the old ghosts in the previous bones dist into the new zig bones file? Seems potentially fine, although I don't know if those ghosts have issues with e.g. removed spells that make them much less effective 09:13:42 <12ge0ff> Could add other extended ghosts into zigs, like players who splatted in Tomb/Hells/Pan. So the ghosts are not just the same mega-mages with Shatter, Fire Storm, and other level 8-9 spells. 09:17:24 <09gammafunk> I'm a bit hazy on how zig ghost collection is done, but I want to say that something like this already happen 09:17:44 <09gammafunk> advil noticed this same issue back when ghosts vaults and the permastore idea were being made 09:19:32 <09gammafunk> hrm, looking at the code, it seems not 09:20:06 <09gammafunk> the permastore size is kind of surprisingly small 09:20:52 <09gammafunk> 10 for individual levels and 20 for branch-wide files 09:26:58 <04dracoomega> Actually, that small number seems to have been for automatically adding to the permastore from live games. I think the only limit when making it manually here is the limit of ghosts per file, which is 127 09:27:12 <04dracoomega> (Which I am hitting for most files) 09:27:38 <04dracoomega> And yes, on reflection, I think I am going to merge some other extended files into the Zig one, rather than old versions of Zigs. Probably still get a reasonable spread of extended-calibre ghosts 09:29:22 <04dracoomega> (Though they're already not as uniformly 'max level blaster spell' as one might think, since it is of course people who died in Zigs) 09:30:15 <04dracoomega> Anyway, I do apparently have to hand-sanitize several thousand names, so I'm hoping I can get that done today, but this entire process has been far more manual than anticipated originally, hence the delays 09:54:34 <04dracoomega> (Frankly, I am not sure the code to write new things to the permastore even does anything these days, since we already ship with a permastore larger than its cap, and to the best of my understanding, ghosts are never deleted from the permastore. But I'm not touching that post-freeze. May take another look at some time in 0.34, though. 10:58:18 03regret-index02 07* 0.34-a0-1-g07266aef2f: Give spell-exclusive allies xp values 10(20 hours ago, 16 files, 21+ 10-) 13https://github.com/crawl/crawl/commit/07266aef2fe0 10:58:18 03regret-index02 07* 0.34-a0-2-g3f916d2844: Sack of Spiders usable / inert tiles 10(3 minutes ago, 4 files, 3+ 1-) 13https://github.com/crawl/crawl/commit/3f916d284407 11:27:02 <09gammafunk> advil took an interesting approach where he kept many player names as-is but did randomization on anything offensive and I'm not sure if there was some other criteria involved. It's neat to be able to see some names that are clearly from real players. Do you intend to randomize all permastore names? 11:27:57 <09gammafunk> I'm trying to remember what was done 7 years ago though 11:28:58 <04dracoomega> Code-wise, it looks like ghosts saved to the permastore online (automatically) were randomized to use the same namegen as pan lord names, while a bunch of other ones were probably manually merged and clearly have played names. I was just going to go through and delete anything overly questionable, since there's a huge number in most of these files anyway 11:29:30 <04dracoomega> (And only a tiny percentage need removal) 11:29:58 <04dracoomega> I actually don't think any of these tools support renaming in a bones file that is already saved, as far as I could see 11:30:00 <04dracoomega> Just removing 11:42:32 <09gammafunk> I see, so no new permastore ghost (in terms of being a ghost we're not shipping in the files) games will ever get an original name. I guess that's what the normal (non-permastore) files are for, though 12:59:57 <04dracoomega> Hmmm... there are apparently actually incredibly few ghosts for Crypt also (which I'd missed on the first pass since I apparently forgot to put Crypt in the script at all >.>) 13:00:12 <04dracoomega> The only other branch besides Zigs where it might be overly thin at the moment 13:00:16 <04dracoomega> (At just 23) 13:00:28 <04dracoomega> Kind of impressed that it was only 23 across all servers 13:01:22 <04dracoomega> (Debating if I should also merge in Depths to Crypt too or something) 13:16:25 -!- adept7 is now known as adept 13:33:20 <09gammafunk> confirmation that crypt is just too easy! 13:34:16 <09gammafunk> I think we're still missing cpo and lld as for as bones files, at this point, unless mcursa did DM you the cpo ones. dplusplus gave me a thumbs up to setting up 0.33 on ldd but no word on the bones files. But both servers should have smaller numbers of ghosts anyhow 13:36:11 <09gammafunk> !lg * current crypt !boring 13:36:16 <04Cerebot> 554. caggimo the Necromancer (L23 CoRe of Hepliaklqana), slain by an ancient champion (a +0 battleaxe of draining) on Crypt:2 on 2025-04-27 20:14:10, with 397865 points after 57546 turns and 3:25:41. 13:36:50 <09gammafunk> hrm, what even is the bones file chance; with 554 in 0.32 + trunk you think there'd be more than 23 ghosts 13:36:58 <04dracoomega> Well, they are removed when used 13:37:12 <04dracoomega> So if someone enters a Crypt Necropolis and then leaves again (whether they kill them or not, I believe) 13:37:42 <09gammafunk> there's nothing that looks at whether the ghost was actually killed? 13:37:50 <04dracoomega> I think not 13:37:56 <09gammafunk> seems unfortunate 13:38:00 <04dracoomega> At least that is how roaming player ghosts used to work 13:38:09 <04dracoomega> Really, there are a bunch of things about this whole system that are pretty imperfect. This is sort of a 'non-representational slice' due to the moment in time that is captured (and the relatively small file cap) 13:39:17 <04dracoomega> Like, this is better than what we have now (centaur fire elementalists with throw flame!), but I can see trying to devise a better way to do this at some future point that actually builds a large enough cache to have a more diverse representation of backgrounds and such 13:39:37 <04dracoomega> (And isn't nearly so manual >.>) 13:40:00 <04dracoomega> I'm finally almost done, though. (I did get CPO bones, by the way) 13:40:20 <09gammafunk> hrm 13:40:37 <04dracoomega> I merged Pan into Zig and Depths into Crypt so those files weren't so tiny and have manually sanitised every resulting file for inappropriate names 13:40:57 <09gammafunk> in end_game(): cpp // Never generate bones files of wizard or tutorial characters -- bwr if (!non_death && !crawl_state.game_is_tutorial() && !you.wizard) save_ghosts(ghost_demon::find_ghosts()); which seems to imply for a game that wasn't a win, live ghosts are saved when the game ends 13:41:19 <09gammafunk> (as long as the game wasn't tutorial or wizard mode) 13:41:19 <04dracoomega> Are portals still 'live'? 13:41:31 <09gammafunk> ah 13:41:33 <04dracoomega> I feel like the map is destroyed when you leave 13:41:39 <09gammafunk> so this is a new problem we have then 13:41:46 <04dracoomega> (I guess I can't be confident what this does to bones) 13:42:04 <09gammafunk> that wouldn't have been as much of a problem with ghost vaults (I guess still an issue for zig ghosts though) 13:43:14 <09gammafunk> hrm, is this really somehow all for just the current level? 13:44:13 <09gammafunk> this looks like it's only dealing with the player's current level and for monsters transitied to the abyss 13:44:28 <09gammafunk> but I guess I'd have to sit down and read it a lot more carefully to see all that's happening 13:46:05 <09gammafunk> ok, great. Maybe we'll never get LLD, but that's literally the least-used server. If dplusplus does get back to me with the files, I can forward them 14:27:08 ratatoskos (L20 DjVM) ASSERT(_valid()) in 'ray.cc' at line 107 failed. (Vaults:1) 14:37:17 03DracoOmega02 07* 0.34-a0-3-g68670cc5b6: Update ghost permastore 10(2 minutes ago, 34 files, 0+ 0-) 13https://github.com/crawl/crawl/commit/68670cc5b658 14:37:18 03DracoOmega02 07[stone_soup-0.33] * 0.33-b1-2-g9666a18ba6: Update ghost permastore 10(2 minutes ago, 34 files, 0+ 0-) 13https://github.com/crawl/crawl/commit/9666a18ba6a8 15:40:07 Unstable branch on underhound.eu updated to: 0.34-a0-3-g68670cc5b6 (34) 16:01:15 <09gammafunk> sweet, we're now mostly ghosted 18:55:32 03regret-index02 07[stone_soup-0.33] * 0.33-b1-3-g47dd9f12d9: Give spell-exclusive allies xp values 10(28 hours ago, 16 files, 21+ 10-) 13https://github.com/crawl/crawl/commit/47dd9f12d9c1 18:55:32 03regret-index02 07[stone_soup-0.33] * 0.33-b1-4-g6f77aabf6c: Sack of Spiders usable / inert tiles 10(8 hours ago, 4 files, 3+ 1-) 13https://github.com/crawl/crawl/commit/6f77aabf6c63 21:11:57 <\amethyst> would it be overly (and anciently) pedantic to argue that, for HC, the plural of "cannon" should be "cannon"? 21:19:13 <\amethyst> also, it might be time for CAO/logs/cheibriados/ not to claim to be temporary, since I don't forsee the s-z.org Chei coming back 21:20:03 <\amethyst> (oh, wait, it's still running? Well, I don't forsee me or anyone else being able to control iot) 21:24:53 <04dracoomega> (I think referring to multiple cannons as cannon will just confuse people. I actually didn't realize the plural was ever not 'cannons', though a cursory investigation suggestions that this was actually a middle point and even Shakespear uses 'cannons') 21:25:12 <04dracoomega> Tsuyoshi is right about cannons being the plural in the 1500's; Google Ngrams doesn't have adequate data before early the 1700's, but we can check Shakespeare, who uses cannon as a regular count noun. So the plural has gone from cannons, to pieces of cannon, to cannon, and back to cannons. 21:25:30 <\amethyst> nice! 21:26:00 <\amethyst> yeah, I didn't actually think "cannons" was wrong, just that "cannon" has more military history energy 21:26:49 <\amethyst> but the inevitable bug reports probably outweigh any "benefit", not that the latter actually exists 21:36:09 <\amethyst> Also, and unrelated, thanks to whoever fixed CAO console yesterday! I was worried enough that I deigned to play webtiles 22:35:56 Unstable branch on crawl.develz.org updated to: 0.34-a0-3-g68670cc5b6 (34) 22:49:17 \amethyst: still running and periodically rejoins the channel (presumably when the machine is rebooted for some reason) 23:00:14 Windows builds of master branch on crawl.develz.org updated to: 0.34-a0-3-g68670cc5b6 23:04:06 SilverEgg (L2 DrSh) ERROR in 'skills.cc' at line 174: ASSERT failed: skill_cost_level of 0 out of range 1 (1) .. MAX_SKILL_COST_LEVEL + 1 (28) (D:1) 23:06:12 !crashlog SilverEgg xl=2 drsh 23:06:13 1. SilverEgg, XL2 DrSh, T:581 (milestone): https://cbro.berotato.org/morgue/SilverEgg/crash-SilverEgg-20250428-060357.txt 23:28:06 Unstable branch on cbro.berotato.org updated to: 0.34-a0-3-g68670cc5b6 (34) 23:56:20 Monster database of master branch on crawl.develz.org updated to: 0.34-a0-3-g68670cc5b6