00:01:31 03Isaac Clancy02 07https://github.com/crawl/crawl/pull/4045 * 0.33-a0-657-g4cb59b7f60: Improve malign gateways targeter 10(5 months ago, 5 files, 101+ 14-) 13https://github.com/crawl/crawl/commit/4cb59b7f6097 00:01:31 03Isaac Clancy02 07https://github.com/crawl/crawl/pull/4045 * 0.33-a0-658-g92d72e468e: Allow malign gateway to be cast in non-compass directions 10(4 months ago, 1 file, 7+ 5-) 13https://github.com/crawl/crawl/commit/92d72e468e1a 00:03:42 03Isaac Clancy02 07https://github.com/crawl/crawl/pull/4045 * 0.33-a0-658-g2d0b93e084: Allow malign gateway to be cast in non-compass directions 10(4 months ago, 1 file, 7+ 9-) 13https://github.com/crawl/crawl/commit/2d0b93e084f6 00:46:30 Monster database of master branch on crawl.develz.org updated to: 0.33-a0-656-g0a0a479413 01:44:08 New branch created: pull/4219 (1 commit) 13https://github.com/crawl/crawl/pull/4219 01:44:10 03Isaac Clancy02 07https://github.com/crawl/crawl/pull/4219 * 0.33-a0-657-g9d62a8fc5a: Improve player tracers 10(27 minutes ago, 26 files, 692+ 485-) 13https://github.com/crawl/crawl/commit/9d62a8fc5a02 01:57:09 03Isaac Clancy02 07https://github.com/crawl/crawl/pull/4219 * 0.33-a0-657-ge1d6dce945: Improve player tracers 10(40 minutes ago, 26 files, 692+ 485-) 13https://github.com/crawl/crawl/commit/e1d6dce9459f 01:58:16 03Isaac Clancy02 07https://github.com/crawl/crawl/pull/4219 * 0.33-a0-657-g08aa2a3056: Improve player tracers 10(41 minutes ago, 26 files, 694+ 487-) 13https://github.com/crawl/crawl/commit/08aa2a3056a0 04:33:11 Experimental (bcrawl) branch on underhound.eu updated to: 0.23-a0-5249-g4a8afe7061 05:37:56 MAN (L27 MiGl) ASSERT(shop) in 'shopping.cc' at line 2016 failed. (Depths:4) 08:12:41 <05icoson> sadly, I think shopping list code probably does fall under that heading 🙂 08:45:16 03advil02 07* 0.33-a0-657-g931650e2db: fix: handle gozag shopping list + abandonment interaction 10(11 minutes ago, 4 files, 33+ 15-) 13https://github.com/crawl/crawl/commit/931650e2db53 09:01:55 <05icoson> because that doesn't restore the daction behavior there's still some sync issues with level maps, and the travel cache still has the abandoned shop items (this probably was broken before, too) 09:03:09 <09gammafunk> would the cleanup-on-abandonment code still have to perform a level excursion in order to do what it needs? 09:03:37 <09gammafunk> I guess the main concern is doing dactions specifically on excursions 09:06:04 <09gammafunk> oh yeah, I see that excursions are still happening with that latest commit 09:06:37 <09gammafunk> obviously need to implement that wishlist item of having space to let two levels be loaded at a time instead of one 09:16:11 <09hellmonk> // Yuck! 09:16:13 <09hellmonk> good comment 09:22:15 <09gammafunk> cio.cc: /* can we say yuck? -- haranp */ dat/database/monspeak.txt:@The_monster@ says, "Orcs smell really bad. Yuck." dat/database/monspeak.txt:@The_monster@ says, "What's that horrible smell? It's like... yuck." dat/database/monspeak.txt:@The_monster@ says, "Yuck, you smell of miasma!" dat/des/variable/mini_monsters.des:# Bad water, a lot of carrion and feeders around, yuck! dat/descript/monsters.txt:breeding experiment. 09:22:15 Yuck. dat/descript/monsters.txt:wizard's breeding experiment. Double yuck! shopping.cc: // Yuck! util/mon-gen.py: # Yuck, we still use negative will for this? 09:58:38 New branch created: shoppricing (1 commit) 13https://github.com/crawl/crawl/tree/shoppricing 09:58:39 03hellmonk02 07[shoppricing] * 0.33-a0-648-gb02f9baa2c: Shop pricing adjustments 10(8 minutes ago, 3 files, 48+ 48-) 13https://github.com/crawl/crawl/commit/b02f9baa2c60 09:58:49 <09hellmonk> @dracoomega this is what I've got on shop pricing so far, let me know what you think 10:33:08 -!- rozlav84 is now known as rozlav8 15:05:29 <04dracoomega> Sure, I'll give it some time when I'm a bit more awake 15:53:49 03gammafunk02 07* 0.33-a0-658-g60181c1710: Make Dissolution amorphous (Colgate) 10(4 minutes ago, 1 file, 1+ 1-) 13https://github.com/crawl/crawl/commit/60181c1710bc 16:00:51 <06pleasingfungus> in my defense re: hand cannon prices, i forgot 16:13:15 <04dracoomega> Broadly agree with most of what I see here. Not sure what I think of the top-end two-handers being literally half the price of the good one-handers. I mean, they're less widely desired, so maybe that makes sense. Feel like they could stand to be a bit higher, though? 200? (This is all so arbitrary, of course) The thing about good weapons (unlike mid-tier weapons) is that you usually only ever need to buy one, so arguably the 16:13:15 price should factor that in? Maybe? Broadly agree with !hw, ?blinking, ?butterflies being in the higher cost tier. Surprised that ?torment also is - maybe just a factor of rarity. Has anyone ever deliberately bought ?torment from a shop, ever, I wonder? (Could probably be a lower tier, imo, though I doubt it matters much.) Not sure if evoker stacking actually means that evokers should cost more? They're already not cheap to pick one up even slightly 16:13:15 early, and the 2nd and beyond is so much less valuable. (The bigger outstanding issue is natural generation rate, imo) Wow, chain cost less than 1/5th the cost of plate? It's like... only slightly worse (but I guess less exciting to buy in the average case.) But I still wonder if the costs should be a little closer there? I also wonder if storm scales should be slightly cheaper? I think shadow scales definitely should be, since it feels like 16:13:15 that armour is almost never worn if not found as some ridiculous randart. (Of course, the way these base prices interact with plusses and brand and such is complex and I haven't looked at that much at all) 16:13:15 <04dracoomega> I don't think we need to div_rand_round anything of this nature - that's for smooth-scaling of things the player has any influence over. I understand the logic of not making antique shop prices scale with depth, but I feel that maybe they still should, just at a smaller rate than before? (And I wonder if regular shops should scale more with depth than even your change.) ...actually, I wonder if a non-linear depth scaling makes any 16:13:15 more sense? A shop on D:2 and D:8 are still kind of in the same 'tier' and one stays at a similar absdepth for a lot of midgame and is the only slightly higher after that, despite having a lot more gold. (I probably would actually put a lower bound on the depth formula randomization, also - no need for it to be possible for a U:5 shop to charge the same prices as D:1, imo.) 16:17:52 <06pleasingfungus> i think i've bought torment for ID and as undead 16:18:04 <06pleasingfungus> not sure i ever remembered to use it, in the latter case 16:18:06 <06pleasingfungus> but i did consider it 16:18:23 <04dracoomega> The first case is basically "Buying with the intent not to use" which is also kind of a sad statement 😛 16:18:28 <04dracoomega> (Not an incorrect one, mind you) 16:19:19 <04dracoomega> You know, on reflection, I wonder if actually makes sense for consumable cost to scale more with depth than equipment cost? 16:20:56 <09hellmonk> Torment pricing is for sure a rarity thing 16:21:10 <04dracoomega> Like, I think we want at least early consumables to be cheap enough that a early player can actually choose to pick some up (there's real decisions there), but later on you can usually just auto-buy anything you think will be relevant. But the deeper into the game, the more items actually struggle to be better than what you already have, regardless of price? And you're usually not buying multiples of most of those things 16:21:26 <09hellmonk> All of the rare scrolls are in that pricing tier currently (except butterflies, which I think is just an oversight) 16:22:18 <09hellmonk> I have more stuff to add but will wait a few hours until I'm off work 16:22:29 <09gammafunk> wow 16:22:39 <09gammafunk> doing crawl dev at work...a travesty... 16:41:26 Unstable branch on underhound.eu updated to: 0.33-a0-658-g60181c1710 (34) 17:41:05 <02Monkooky> I'll note that buying unID'd things early is real powerful; finding an early potion or scroll shop is a huge power swing 17:41:21 <02Monkooky> so making consumables cheaper early is probably not a good idea 17:53:31 <09hellmonk> Most of the individual price changes are indeed sort of chasing the 'meta' (which comes with some risks of its own). Could see adjusting 2handers up or bringing the top 1handers back down slightly, I'm not particularly sold on the specifics. Agree re: evokers, will revert that. Can see going some lower on the higher end dragon scales as well, though I think it's fine that the special rare armours that drop from monsters kind of suck to 17:53:32 buy. Pluses are additive, as are most armour egos and randart special stuff iirc, but weapon brand is multiplicative. Will make a separate post about shop prices in a few minutes. 17:55:02 <04dracoomega> No, this was mostly aimed in the opposite direction. Making them more expensive later and not changing early stuff much at all. 17:55:26 <04dracoomega> Early shop balance mostly feels pretty good to me, with actual interesting decisions being made 18:09:27 <09hellmonk> re: depth scaling, I am really a bit unsure what the best way to handle it is. In some sense it feels like other shop knobs (item quality, how many shops you see) should be "enough" to not need this hidden scaling factor but I think it's probably necessary given crawl's current systems. I do wonder about attacking the supply side and just not generating as many potion and scroll shops later, but then players have even more useless gold 18:09:28 I suppose (was also vaguely worried about having to backtrack for consumables more but I think variable greed means you almost always do this anyway). Changing the scaling to just random2(level) felt very noticeable in my brief testing, +0 flaming flail was like 500 gold in depths 2. Could have been rng though. WRT antique versus regular shops, I think the current system works alright for stuff that is actually unided and is very bizarre for stuff that 18:09:28 isn't unided but shows up in the shop anyway (unrands, consumables the player has seen) which is much more expensive than if the item showed up in a normal shop. I actually wonder if it might be better to use identical greed formulas and raise the added cost for unidentified items. The current numbers will make antique shops still usually more greedy than regular shops due to that much larger rand I think. 18:11:01 <09hellmonk> I guess my thought process is that we want to use the greed formula to limit the number of consumables the player can buy lategame somewhat, and we want unidentified items from antique shops to be somewhat discounted because they're a gamble but identified items in antique shops probably do not need to be more expensive than normal. 18:14:02 <09hellmonk> in some sense it might even be good to make a distinction between consumables (need to scale in cost because of the amount of gold crawl generates, they stay ~equally useful all game, and don't want to lose early shop choices) and equipment (generally becomes less significant of an upgrade as the game progresses and more likely to be totally useless because a better item has generated, so less reason to scale the cost). Unfortunately I 18:14:07 can't think of a good way to only make consumable costs increase that is not very stupid. 18:16:17 <04dracoomega> I guess my thought process is that we want to use the greed formula to limit the number of consumables the player can buy lategame somewhat, and we want unidentified items from antique shops to be somewhat discounted because they're a gamble but identified items in antique shops probably do not need to be more expensive than normal. I can agree with this in general. There's already a lot of item-specific code for prices. Is there 18:16:17 any reason that different item types couldn't interpret greed differently, if that is basically the effect we want? (I guess this maybe has somewhat weird effects with shops that specifically try to use unusual greed, like Platinum Reserve or Big Baal Hell's?) 18:16:52 <04dracoomega> Platinum Reserve seeling an unrand at mildly-above-cost and a potion at 10,000x markup >.> 18:17:06 <09hellmonk> I suppose it's possible to do that (but it feels quite hacky) 18:17:57 <04dracoomega> I mean, is it really hacky if what we want is for certain types of items to scale differently than other item types? 18:18:20 <04dracoomega> (Except for vault stuff) 18:18:28 <09hellmonk> yeah, maybe it's fine 18:19:52 <09hellmonk> I'll poke at it some more this weekend 20:12:49 03Cgettys02 07https://github.com/crawl/crawl/pull/4217 * 0.33-a0-657-g4d30fc8c4d: Avoid depending on deprecated, terrible GLU library by dynamically loading glGenerateMipmap where available. 10(25 hours ago, 8 files, 100+ 17-) 13https://github.com/crawl/crawl/commit/4d30fc8c4df7 20:17:10 03Cgettys02 07https://github.com/crawl/crawl/pull/4218 * 0.33-a0-657-g31d9fb68a5: Get coverage to warn instead of error 10(22 hours ago, 1 file, 11+ 2-) 13https://github.com/crawl/crawl/commit/31d9fb68a51d 20:40:05 03Cgettys02 07https://github.com/crawl/crawl/pull/4217 * 0.33-a0-659-g092ec479ba: Avoid depending on deprecated GLU library by dynamically loading OpenGL 3+ function glGenerateMipmap where available. If it's not available, Where it's not, we just render without mipmapping. 10(25 hours ago, 8 files, 100+ 17-) 13https://github.com/crawl/crawl/commit/092ec479bae3 20:55:05 Leverkuhn (L27 NaCA) ERROR in 'beam.cc' at line 745: ASSERT failed: flavour of 0 out of range BEAM_NONE + 1 (1) .. BEAM_FIRST_PSEUDO (100) (Pan) 20:56:24 <09gammafunk> that is certainly an interesting assert message 20:56:58 <09gammafunk> sadly on CAO so we're not going to get much in terms more than stack frames 20:57:04 <09gammafunk> !crashlog leverkuhn 20:57:06 <04Cerebot> 3. Leverkuhn, XL27 NaCA, T:116574 (milestone): https://crawl.akrasiac.org/rawdata/Leverkuhn/crash-Leverkuhn-20250111-035504.txt 20:57:10 Leverkuhn (L27 NaCA) ERROR in 'beam.cc' at line 745: ASSERT failed: flavour of 0 out of range BEAM_NONE + 1 (1) .. BEAM_FIRST_PSEUDO (100) (Pan) 20:57:24 RE: https://github.com/crawl/crawl/actions/runs/12720883662/job/35463199600?pr=4217#annotation:11:1945ld: warning: ignoring file '/Users/runner/work/crawl/crawl/crawl-ref/source/species.o': found architecture 'arm64', required architecture 'x86_64' 20:57:29 ld: warning: ignoring file '/Users/runner/work/crawl/crawl/crawl-ref/source/species.o': found architecture 'arm64', required architecture 'x86_64' 20:57:34 this is why the linker is failing 20:57:47 Ijust don't know why it's ending up in that situation in the firs tplace 20:58:30 <09gammafunk> right, there is something weird that happens (sometimes?) on CI builds for macos linking for the universal binary 20:58:42 <09gammafunk> can I just try to re-run this task I wonder 20:59:01 <09gammafunk> just did so, we'll see if it happens again 20:59:14 Yeah, it wouldn't surprise m e if it's a fun race condition 20:59:33 <09gammafunk> there are a lot of transient CI problems/errors 20:59:49 universal anything is kinda nightmarish 20:59:51 <09gammafunk> that seem to boil down to certain packages/dependencies sometimes failing to be set up correctly or installed at all 20:59:59 wheni t comes to native code 21:00:05 <09gammafunk> https://github.com/crawl/crawl/actions/runs/12720883662/job/35463376496?pr=4217 21:00:33 linkers are the ugly forgotten but critical bits of teh build process, and they're kinda nightmarish 21:00:55 add trying to smush two different sets of object files together... 21:01:06 at least you're not trying to build ARM64EC, I guess 21:01:11 <09gammafunk> it succeeded 21:01:15 Hurray 21:01:20 <09gammafunk> so yeah just one of those fun transient errors 21:02:05 Could be worse 21:02:21 Imagine your fun transient errors often take 30 minutes or an hour to retry 21:02:25 instead of 2 minutes 21:02:30 I don't have to imagine :D 21:02:59 it's still complaining above 1 of the checks getting canceled 21:03:11 can you try retrying macos tiles? think that's why it's grumpy 21:03:25 https://github.com/crawl/crawl/actions/runs/12720883662/job/35463376376?pr=4217 21:03:51 The job was canceled because "console-universal" failed. 21:05:58 <09gammafunk> you're far from the first person working on this project to have to deal with slow compilation times :p 21:06:11 I wasn't complaining about compilation times 21:06:36 I was complaining about tests :) 21:06:46 Welcome to enterprise development :) 21:06:53 <09gammafunk> or slow anything related to project infrastructure! 21:07:10 Oh, sure, but your infrastructure isn't as bad as you think :D 21:07:15 <09gammafunk> this project's been running on resources strung together by a mix of devs and admins since 2006 21:07:45 <09gammafunk> trust me, one of the two of us knows what our infrastructure is, and it's definitely got plenty of problems 21:07:54 I never said it didn't :P 21:08:06 <09gammafunk> I would not call it close to enterprise level by any stretch 21:08:10 I'm saying something very different 21:08:24 Flaky tests and slower builds are not uncmmon in large enterprises 21:08:45 <09gammafunk> well I have great news, both are extremely common in this project as well! 21:08:54 Feels just like home 21:09:04 Seriously though your CI is faster than I'm used to :D 21:09:30 And it's less forgiveable when there's an enterprise budget paying for it :D 21:09:36 <09gammafunk> yeah it's not "our" CI or anything; it's 100% built by github and we're just using it. Indeed github has fast servers 21:09:39 <03implojin> crawl is run on shoelaces and hope 21:09:44 <09gammafunk> word 21:10:08 We have more shoelaces, and sometimes less hope. But also way more code 21:10:19 Seriously, I removed 20000 lines of code this week :D 21:10:23 <09gammafunk> we have "endless" CI problems but it's also quite useful 21:12:39 We'll see about that, I guess 21:13:11 <09gammafunk> see about what? 21:13:56 <09gammafunk> anyhow that macos tiles job did rerun successfully, so sadly those types of macos linking failures are just another CI error that has to be ignored (there are a number of those) 21:16:08 Whether it's endless :D 21:16:48 SEe, you guys have something like 500k lines of C++ if cloc is to be believed 21:17:21 Not sure if that's including a lot of contrib stuff, I tried to exclude it 21:17:47 If I go fix 10000 lines over a few months, that's actually wow, noticeable 21:18:13 it's 2% 21:18:22 If I fix one of the CI targets to be reliable, it's like 10% 21:18:44 if I do that at my day job... it does add up, but there's an order magnitude more to fix 21:19:00 :D 21:20:16 <09gammafunk> I mean, when I say we have endless CI problems, that is speaking generally and from experience 21:20:36 <09gammafunk> our CI problems do not consist entirely of this one particular macos problem 21:20:55 <09gammafunk> so yes you'll continue to see CI breakage and failures even if this one resolves itself 21:21:15 Sorry, I'm tired and being unclear as hell 21:21:19 <09gammafunk> that's been our experience with github CI from the beginning (and also with github's Travis system that preceeded their CI and github actions) 21:21:50 What i was (somewhat) jokingly saying is "we'll see if your CI problems are actually more endless than my stubbornness to fix CI problems" 21:22:00 <09gammafunk> yeah I'm just emphasizing that this project has a lot of prior history and experience going back almost two decades 21:22:11 <09gammafunk> so there's lots of ups and downs, like any long-lived project 21:22:34 Of course 21:22:35 <09gammafunk> we're not an enterprise system, we're a "mid-sized" video game, but there's a lot of broken stuff in all areas 21:22:51 not trying to minimze your guys' experience 21:24:31 <09gammafunk> what you'll find is that, despite this project being "only" e.g. 500k lines of C++ (and a few thousand lines of lua, and thousands of lines of python and javascript etc etc) is that despite it being smaller, there are just a lot of thorny fundamentally hard to resolve problems that sometimes one can only sort of poke at a bit 21:24:45 Sounds like my day job :D 21:24:46 <09gammafunk> so it's just not the case that you run out of broken things to work on or anything 21:24:55 Oh, I didn't expect to ever get to 100% 21:25:05 You never run out of broken things to work on 21:25:15 but maybe if we'll lucky I'll get us to the point where CI is reliabl-ish 21:26:42 <09gammafunk> well, the transient CI problems are a mix of fundamentally github-side things and yeah some probably that we can improve 21:27:17 <09gammafunk> like a bunch of times they literally fail to install debian packages from their CI repos as part of setup, and when that happens nothing you can really do would prevent those build failures 21:27:36 <09gammafunk> but improving CI is definitely good and it's a system we have to keep poking at and improving 21:28:24 wow, this is pretty busted, yeah 21:28:42 Can't get version information: `git describe` failed (no git, no repository, or shallow clone), and util/release_ver doesn't exist. 21:28:47 fun :) 22:28:11 I understand I'm hitting the git describe failure, kinda 22:28:44 what I don't understand is how @implojin 's branch manages to get the pstream set up right remotely 22:40:46 ah, git push origin --tags 23:35:14 Unstable branch on crawl.develz.org updated to: 0.33-a0-658-g60181c1710 (34) 23:35:32 -!- Changesite1 is now known as Changesite 23:57:52 Windows builds of master branch on crawl.develz.org updated to: 0.33-a0-658-g60181c1710