00:41:40 Fork (bcrawl) on crawl.kelbi.org updated to: 0.23-a0-4907-g988a294422 03:17:09 Route97 (L6 MiBe) ASSERT(desc_tgt) in 'item-use.cc' at line 767 failed. (D:4) 03:23:59 Experimental (bcrawl) branch on underhound.eu updated to: 0.23-a0-4907-g988a294422 04:31:16 New branch created: pull/3076 (1 commit) 13https://github.com/crawl/crawl/pull/3076 04:31:16 03Medrano8302 07https://github.com/crawl/crawl/pull/3076 * 0.30-a0-1009-g9bbf95c8e0: Include a reference to fungus in the combat tutorial 10(23 minutes ago, 2 files, 34+ 25-) 13https://github.com/crawl/crawl/commit/9bbf95c8e0bb 05:30:26 Unstable branch on crawl.akrasiac.org updated to: 0.30-a0-1008-g30c9307 (34) 07:36:59 03Medrano8302 07https://github.com/crawl/crawl/pull/3076 * 0.30-a0-1009-g3b3d2b9731: Include a reference to fungus in the combat tutorial 10(3 hours ago, 2 files, 34+ 25-) 13https://github.com/crawl/crawl/commit/3b3d2b97311a 07:51:19 regis (L9 DsGl) ERROR in 'tileweb.cc' at line 234: Socket write error: Resource temporarily unavailable (D (Sprint)) 07:51:42 03Medrano8302 07* 0.30-a0-979-g8c6280c012: Fix hardcoded font paths in Debian packages 10(5 days ago, 2 files, 3+ 3-) 13https://github.com/crawl/crawl/commit/8c6280c01225 07:51:42 03Medrano8302 07* 0.30-a0-980-g28e3bdbff4: Fix AppImage builds 10(5 days ago, 1 file, 2+ 0-) 13https://github.com/crawl/crawl/commit/28e3bdbff4b8 07:51:42 03Medrano8302 07* 0.30-a0-981-gf56873caf1: Fix GitHub Actions output parameters 10(5 days ago, 1 file, 3+ 3-) 13https://github.com/crawl/crawl/commit/f56873caf1fd 07:51:42 03Medrano8302 {GitHub} 07* 0.30-a0-1012-g6f0cfe7724: Merge pull request #3067 from robertxgray/hotfix/debian_fonts 10(16 seconds ago, 0 files, 0+ 0-) 13https://github.com/crawl/crawl/commit/6f0cfe772454 07:52:40 <10PleasingFungus> hooray, a merge! 07:53:16 <10PleasingFungus> @robertxgray minor: commit history tends to be a little cleaner and easier to follow if you rebase rather than merging (in most cases) 07:55:05 <03robertxgray> That was my intention, I must have done a mistake 07:55:34 <10PleasingFungus> no worries 🙂 08:00:38 <03robertxgray> I guess I shoudn't have added a comment after selecting "Rebase and merge", sorry 08:03:49 Unstable branch on crawl.kelbi.org updated to: 0.30-a0-1012-g6f0cfe7724 (34) 08:39:06 Since CAO is in pretty bad shape, I could create an azure vm in that region to host a server for the tournament, if there would be a need for something like that. 09:11:50 <10PleasingFungus> thank you! hopefully that won’t be needed 09:18:59 agree regarding scorch nerf. It could be really strong to blindly spam early to mid game. 09:22:02 <06advil> rax has been checking in to the cao situation 09:22:04 <08nicolae> i'll put it on my thinky-think to-do list 09:23:36 <06advil> adding a new official server is a bit non-trivial, we may or may not want to do it sometime soon but it's not necessarily volunteering for a quick and easy task; if CAO in particular has to be replaced at some point I would prefer to administer it myself 09:24:24 <06advil> it is possible we're low on server capacity relative to demand? 09:24:28 sentei does have the necessary know-how, FWIW 09:25:07 How much memory/storage space is usually needed? 09:25:12 <10PleasingFungus> if we add a new official server, would it be possible to import the login db from an existing server? i assume we hash & salt pws 09:25:27 <06advil> yes, we did that with CBRO->CBR2 09:25:35 <06advil> though the cao login db is gigantic and old 09:25:54 <06advil> but throwing open the gates also causes problems 09:26:28 <10PleasingFungus> seems like we should import a db from somewhere - really using any server’s login db would be much better than a fresh slate 09:26:40 <10PleasingFungus> in terms of minimizing potential for abuse 09:26:51 <10PleasingFungus> how gigantic is gigantic? 09:27:23 <06advil> !lg * cao s=uniq(name) 09:27:25 <04Cerebot> Unknown function: uniq 09:27:30 <10PleasingFungus> cdist? 09:27:47 <10PleasingFungus> !lg * cao x=cdist(name) 09:28:13 <04Cerebot> 4430342 games for * (cao): cdist(name)=59427 09:28:17 <06advil> I can check the db too I'm just not connected at the moment 09:28:18 <10PleasingFungus> oh 60k 09:28:32 <10PleasingFungus> presumably slightly higher from people who never finished a game 09:29:26 <10PleasingFungus> if you give 1kb per user, that’s 60mb? who cares 09:29:33 <10PleasingFungus> postage stamp db 09:31:17 <06advil> I don't mean big in terms of raw size, but big in terms of webtiles server interacting with an sqlite3 db 09:31:21 <10PleasingFungus> (would be surprised if u/n+pw is 1kb avg) 09:31:23 <10PleasingFungus> hm 09:31:32 <10PleasingFungus> wdym? slow? 09:31:39 <06advil> it's not the most efficient thing 09:31:46 <06advil> though I did some tweaks last year sometime 09:32:29 <06advil> !git e1e7d3de80dd 09:32:30 <04Cerebot> advil * 0.30-a0-353-ge1e7d3de80: fix: add a unique index for dglusers.username (5 months ago, 1 file, 25+ 3-) https://github.com/crawl/crawl/commit/e1e7d3de80dd 09:32:34 <06advil> maybe it's fine after that commit? 09:32:40 <06advil> it wasn't fine before it 09:32:56 <06advil> ah, also that commit msg claims 100k rows 09:32:57 <10PleasingFungus> interesting 09:33:38 <10PleasingFungus> my intuition is that this should be very fast, but intuition and performance are not good friends 09:33:43 <06advil> !git dc488dc59d06d 09:33:43 <04Cerebot> advil * 0.30-a0-349-gdc488dc59d: refactor: rewrite webtiles db code (5 months ago, 4 files, 486+ 310-) https://github.com/crawl/crawl/commit/dc488dc59d06 09:33:46 <06advil> also that commit 09:35:13 <10PleasingFungus> oof 09:35:22 <10PleasingFungus> yeah opening a new connection each time would be 09:35:24 <10PleasingFungus> incorrect 09:36:47 Also to clarify a bit regarding the offer to spin up a vm for the tourney, I did mean that I intend it to be temporary for the duration of the tourney to reduce load, not to offer to be the new permanent server. 09:37:38 I did do some editing of the docker webtiles scripts to hopefully make the spin up process for the webtiles faster 09:37:46 to host the fork server 09:43:13 <06advil> I think it's entirely plausible in 2023 to design a user db that scales well and can handle 100k accounts with absolutely no problem, and it's possible that I have retroactively managed this, but from profiling before those changes even locally on a fast ssd-based device it wasn't amazing; I did some profiling after and iirc it looked good but I can't promise much about how it scales further 09:43:34 <06advil> so your intuition is probably right about anything like this one would devise from scratch nowadays 09:43:59 <06advil> I guess a lingering question I would have is if sqlite3 is really the right way for this sort of scaling? 09:44:17 <06advil> e.g. putting the dream sso system in sqlite3 sounds like an iffy design choice 09:44:30 <10PleasingFungus> my work uses postgres 09:44:42 <10PleasingFungus> we’ve been generally happy with it 09:44:52 <06advil> yeah either postgres or mysql ought to be better, potentially a lot better (so says my intuition0 09:45:29 <06advil> more complex to set up though 09:45:57 <06advil> sentei: thanks! we'll see what happens next with cao 09:46:00 <10PleasingFungus> we have some 5-20TB postgres DBs that run pretty well (and various much smaller ones) 09:46:19 <10PleasingFungus> i can help out with setup if you want to experiment 09:47:26 <10PleasingFungus> JOKE: mongodb 09:47:34 <06advil> really the "urgent" need for db optimization is in scoring 😉 09:48:03 <06advil> though step 1 might be to move that off of cao 09:48:12 <10PleasingFungus> seems reasonable 09:48:23 <10PleasingFungus> microservices…….. 09:48:24 <06advil> (rax was talking about replacing a disk in the last email so we'll see) 09:49:07 <10PleasingFungus> did sound like a dying disk 09:49:26 <10PleasingFungus> from your disk stats reports 10:00:43 Csilarion (L12 HOFi) Crash caused by signal #6: Aborted (D:13) 12:20:38 <08wormsofcan> when are we adding the seven orbs of the dragon king unrands 12:23:29 <10PleasingFungus> nicolae: re unrand orbs, my sources of inspiration are: 1. exaggerate an orb ego. what does an orb of super-light look like, or an orb of super-mayhem. 2. what takes good advantage of the 'orb' item type? what's something that's cool as a 'blocks you from using a shield or 2-hander' tradeoff? 3. puns (orb of destruction, orb of fire, sphere of battle...) 12:24:22 <08nicolae> imo super-mayhem would be more like what i originally suggested: whenever anything dies in your LOS, everything in LOS has a chance to get frenzied, so it would escalate 12:24:30 <10PleasingFungus> sure 12:25:40 <08nicolae> fr the Lament Configuration as an orb (but it's a cube) 12:33:20 <08wormsofcan> orb that summons frenzied demons when you kill things 12:38:53 <10PleasingFungus> worrying 12:39:32 sounds like a good way to become by-catch 12:53:06 69420guy (L2 GrFi) Crash caused by signal #6: Aborted (D (Sprint)) 13:07:49 <08nicolae> in the words of the ancients: Shoulda Thought About That Before You Did It 13:28:57 <13Sastreii> For real! We already have majin boo 13:33:11 <09hellmonk> Next April fool's replace parghit with perfect cell 13:39:04 03PleasingFungus02 07* 0.30-a0-1013-gde262b677f: Fix plasma beam targeter display 10(3 hours ago, 1 file, 1+ 1-) 13https://github.com/crawl/crawl/commit/de262b677f82 13:39:04 03PleasingFungus02 07* 0.30-a0-1014-g3371b13838: Show AFF_YES for plasma beams with certain targets 10(28 seconds ago, 3 files, 42+ 36-) 13https://github.com/crawl/crawl/commit/3371b1383805 13:49:36 <06advil> yes, apparently some raid thing had a bunch of errors (I'm butchering the description) 13:56:20 Unstable branch on crawl.kelbi.org updated to: 0.30-a0-1014-g3371b13838 (34) 14:12:46 hihot (L14 HOFi) Crash caused by signal #6: Aborted (Gauntlet) 15:01:30 <08nicolae> okay, so i was thinking about 2 here. what's one group who would likely be using a shield or 2-hander? weapons users. your MiFis and whatnot. what might they want in exchange for SH or Hit Things Harder? i thought of a few ideas but they were all just variations on Deal More Damage But With Extra Steps or Take Less Damage But With Extra Steps. what might a hit-things build want besides those? what about something that would make it 15:01:30 easier to cast in heavy armor, but such that it wouldn't be useful to your average deep elf. 🤔 how about: The Orb Of The Muscle Wizard. use your str in place of your int when casting. maybe some other magic effects to help you cast better. 15:05:37 <10PleasingFungus> makes me very nervous 15:05:48 <10PleasingFungus> anything that lets you totally ignore one stat is a very strong effect 15:06:34 <10PleasingFungus> seems easy to imagine someone with this orb just being a vastly stronger character than a wizard who bothered investing in int 15:07:41 <13Sastreii> i finally did it. randart version of all the dragon scales. 15:07:42 <13Sastreii> https://cdn.discordapp.com/attachments/747522859361894521/1093295926388473896/Schermata_2023-04-06_alle_00.00.17.png 15:07:42 <13Sastreii> https://cdn.discordapp.com/attachments/747522859361894521/1093295926635925584/dragon_armor.zip 15:08:03 <13Sastreii> (top: current; bottom: randart version) 15:15:41 <08nicolae> hmmm. true. what if it was some kind of casting bonus based on your armor penalty, then 15:15:59 <08nicolae> hell yeah 15:17:09 <10PleasingFungus> wow! 15:18:05 converts your armour malus to a casting bonus… meaning you also lose the armour bonus while wearing the orb 15:19:39 (but still suffer its downsides, so you can't just put on ocpa with the orb to become a master blaster in effective leather) 15:22:05 <10PleasingFungus> nicolae: some kind of bonus based on evp could work. would take careful design 15:22:13 <10PleasingFungus> geekosaur: I don't really understand, sorry 15:22:23 <08nicolae> if there's one thing i'm known for, it's careful design 15:23:47 yeh, I didn't really think it through at first. the idea would be basically what you just said, turning some amount of your aevp into a casting bonus. but you still need enough str to wear the armour and you still have an evasion penalty, so it's not just "heavy armour = free bonus" 15:24:08 which amounts to that careful design you mentioned 15:24:42 Unstable branch on underhound.eu updated to: 0.30-a0-1014-g3371b13838 (34) 15:24:52 <10PleasingFungus> can imagine a chance at mp refund based on the armour penalty, but that's a little closer to orb of energy than i'd like. can imagine a power bonus, but that's maybe too close to orange crystal plate 15:25:09 <10PleasingFungus> probably there's some other idea 15:31:42 <08nicolae> reduction of failure chance? 15:31:54 <08nicolae> maybe mp regen 15:32:56 03PleasingFungus02 07* 0.30-a0-1015-gecba963dcd: Randart dragon scale tiles (Sastreii) 10(10 minutes ago, 12 files, 15+ 0-) 13https://github.com/crawl/crawl/commit/ecba963dcd14 15:33:07 <10PleasingFungus> i thought about reduced failure chance, but that feels a little weird, since increased failure chance is the effect of heavy armour. maybe it'd be fine, idk 15:33:23 <10PleasingFungus> mp regen is mp refund but with extra steps 😛 15:33:33 <08nicolae> actually, mp refund is mp regen with extra steps 15:33:36 <10PleasingFungus> hm 15:34:08 <08nicolae> just split the difference and give it "Int + X" where X varies based on your armor penalty 15:34:27 <08nicolae> (split the difference between the two ideas i had, that is) 16:02:29 <08nicolae> anyway. i will ponder the orb some more 16:03:01 <08nicolae> ultra version of orb of guile. nobody in LOS gets any willpower whatsoever. "immune to hexes"? not today motherfucker 16:10:54 Unstable branch on crawl.kelbi.org updated to: 0.30-a0-1015-gecba963dcd (34) 16:15:08 <08nicolae> part of the difficulty is that orbs don't have a consistent mechanical identity beyond "they're not shields" 🤔 16:31:09 <10PleasingFungus> they're an upside and a downside 16:31:16 <10PleasingFungus> (devs LOVE downsides. they're addicted to downsides) 16:31:25 <10PleasingFungus> is this vague? yes 16:31:43 <10PleasingFungus> but it's a stronger identity than any other slot's egos have 😛 16:34:56 <02deltafunk> the upside of being on the moon is no gravity 16:34:59 <02deltafunk> the downside is no air 16:38:22 <08nlavsky> re: unrand orbs, there could be the orb of antigravity that halves your armour's ER 17:03:08 <06advil> I just spent like 20 minutes looking for the older version of https://github.com/crawl/crawl/issues/1497 before realizing it was originally posted in 2020 17:22:47 <05kate> i feel like to some extent unrand orbs could also just do the orb of dispater thing of "cool effect that you should need to give up an equipment slot to use and shouldn't require a bunch of weapon swapping". i could definitely see converting wucad mu and staff of battle to be orbs too for example (also ties into hellmonk's agenda of having enhancer staves be offhand items?) 17:24:31 <05kate> maybe wucad mu as an orb just being straight-up "it's a stronger orb of energy" isn't super exciting but we have plenty of other unrands along those lines and imo getting rid of the WPN_STAFF item entirely is a good goal to have 17:26:25 <05kate> (other obvious pro is that staff of battle as an orb fits the "puns" category perfectly) 17:27:36 <02deltafunk> you want to get rid of WPN_STAFF melee entirely? 17:29:29 <05kate> should make enhancer staves all make you vulnerable to their opposite element (again? did they ever do that or was it only the spell de-enhancer) so that they all have the requisite downside to be converted to orbs... 18:50:38 <10PleasingFungus> yeah staff of battle -> orb was on my list for that exact reason 18:52:30 <10PleasingFungus> gamma: don’t think we need to change any functionality, just make em OBJ_WEAPON with SPWPN_CONJ or w/e. i almost did this when i supported randart staves - it’d be a nice code simplification for all things you wield in hand 1 and hit stuff with to be the same base type 19:13:35 <02deltafunk> ah, I see, thanks 19:24:38 <10PleasingFungus> possibly kate did want to turn all magic staffs into orbs? i’m a little lower on that because it means casters just always end up holding sky weapons 19:25:17 <10PleasingFungus> seems like worse lootfeel 19:25:20 <10PleasingFungus> but idk 21:56:07 03PleasingFungus02 07* 0.30-a0-1016-gfb0d676c02: Fix friendly fire 10(56 seconds ago, 1 file, 2+ 1-) 13https://github.com/crawl/crawl/commit/fb0d676c0221 22:04:52 Unstable branch on crawl.kelbi.org updated to: 0.30-a0-1016-gfb0d676c02 (34) 23:38:30 Unstable branch on cbro.berotato.org updated to: 0.30-a0-1016-gfb0d676c02 (34)