01:42:43 ZaQ (L17 TrDe) ERROR in 'mon-util.cc' at line 3208: bogus mc (no monster data): invalid monster_type 1000 (1000) (Orc:1) 02:52:55 03CrawlOdds02 07* 0.35-a0-464-g6a2dd7a36a: Allow positive mutations while at innate stat-zero (AliasTheSpectator) 10(22 minutes ago, 1 file, 7+ 7-) 13https://github.com/crawl/crawl/commit/6a2dd7a36a93 02:59:48 04Build failed for 08master @ 6a2dd7a3 06https://github.com/crawl/crawl/actions/runs/27059138354 03:28:26 <04dracoomega> Oh, huh. That build failure didn't get announced earlier, but seems related to the stuff I pushed 03:30:39 <04dracoomega> Concerning that mon-gen.py can fail on the MSVC build but not otherwise. Is python somehow parsing the unicode characters differently? 03:31:09 <04dracoomega> Like, on that build test, it seems to think Σ is multiple characters long 03:35:26 Experimental (bcrawl) branch on underhound.eu updated to: 0.23-a0-5261-gd9800d219b 03:36:40 <04dracoomega> Like, it sounds like some kind of encoding issue, but it's just UTF-8 03:52:01 maybe the locale was just wrong? 03:52:25 another project I've been involved with has multiple times gotten mis-provisioned runners with odd problems like that 03:53:07 seems to be happening more often of late; I think GH's hitting a limit somewhere and the backend is randomly failing 03:56:27 03DracoOmega02 07* 0.35-a0-465-gcc6d5d2c9e: Potentially fix a CI issue with the MSCV build 10(46 seconds ago, 1 file, 1+ 1-) 13https://github.com/crawl/crawl/commit/cc6d5d2c9e11 03:57:20 <04dracoomega> (I ran into some other people talking about issues with PyYaml on windows falling back to a system locale thing and not reading utf-8 files properly without being forced, so I thought I'd at least test this) 03:57:51 <04dracoomega> Somewhat awkward given that I don't exactly have a local MSVC environment available 04:02:24 04Build failed for 08master @ cc6d5d2c 06https://github.com/crawl/crawl/actions/runs/27060389445 04:05:26 <04dracoomega> Well dang 04:08:15 <11Odds> I have a windows machine and so presumably could get this build running 04:09:37 <11Odds> Though even then the locale thing could be quite specific to the machine its building on 04:10:49 <04dracoomega> I really, really, really don't want to replace the glyphs in the yaml with codepoint literals. The loss of readability is miserable. 04:11:10 <04dracoomega> (mon-gen.py already theoretically handles converting that when generating mon-data.h) 04:13:26 <04dracoomega> (I also have a windows machine and could presumably eventually get some MSVC build running, but beyond having no experience with it or the process involved, INSTALL.md does not exactly inspire confidence here: This build process is currently unsupported, and unlikely to be straightforward in versions of MSVC besides those explicitly mentioned here. 04:15:50 <11Odds> Probably we just need to specify the encoding to the open call in the python script 04:17:32 <11Odds> Though I really thought python just defaults to utf-8 rather than caring about the environment, so maybe that’s wrong 04:17:51 <04dracoomega> (I mean, I thought the command line switch I added was supposed to force that regardless) 05:11:14 <11Odds> Oh right I think the problem is that gen-all launches the individual gens as subprocesses that don't get the argument 05:13:03 <11Odds> And I think set PYTHONUTF8=1 would make sure they all picked it up from the environment 06:03:47 DO: I'm liking the new cloud glyphs (among many others). Good stuff! :) 06:05:32 We had to change one, it came out quite badly in Bedstead :-/ 10:32:10 New branch created: goodlubria2 (1 commit) 13https://github.com/crawl/crawl/tree/goodlubria2 10:32:10 03hellmonk02 07[goodlubria2] * 0.35-a0-453-g61eb8347c6: Refactor passage of golubria 10(11 minutes ago, 1 file, 32+ 38-) 13https://github.com/crawl/crawl/commit/61eb8347c64b 10:32:10 New branch created: pull/5287 (1 commit) 13https://github.com/crawl/crawl/pull/5287 10:32:10 Branch pull/5287 updated to be equal with goodlubria2: 13https://github.com/crawl/crawl/pull/5287 11:16:16 <02Monkooky> wow old golubria code was stinky 11:44:09 <09gammafunk> ....crawl code?! IN MY REPOSITORY?!?! 12:10:16 <09hellmonk> an interesting thing I discovered while working on this is that a pog portal can spawn on your tile, but only if it's the distant portal being placed there by fuzzing 12:10:40 <09hellmonk> the close portal is not allowed to spawn under you 12:11:10 <09hellmonk> probably some absurd edge case where you can try to save a turn with this knowledge 12:50:48 <09gammafunk> I will program this knowledge into qw to achieve 100% winrates 15:42:40 Unstable branch on underhound.eu updated to: 0.35-a0-465-gcc6d5d2c9e (34) 16:35:22 -!- MakMorn_ is now known as MakMorn 17:31:20 03refracta02 {gammafunk} 07* 0.35-a0-466-gbcb5613262: Limit path macros to dependent objects 10(4 weeks ago, 1 file, 31+ 7-) 13https://github.com/crawl/crawl/commit/bcb56132623f 17:39:51 03DracoOmega02 07* 0.35-a0-467-g0a0f4f6124: Another attempt at fixing MSVC CI issues 10(64 seconds ago, 2 files, 2+ 2-) 13https://github.com/crawl/crawl/commit/0a0f4f6124e2 17:46:23 04Build failed for 08master @ bcb56132 06https://github.com/crawl/crawl/actions/runs/27078074863 18:00:17 <09gammafunk> rip the fix 18:51:46 <04dracoomega> No, that was a delayed response to the commit you pushed. Mine did in fact seem to work. 22:35:45 Unstable branch on crawl.develz.org updated to: 0.35-a0-467-g0a0f4f6124 (34) 22:59:41 Windows builds of master branch on crawl.develz.org updated to: 0.35-a0-467-g0a0f4f6124 23:00:44 <11Odds> Hmmm, I'm a long way down the rabbit hole of moving penance conducts to hard blocks on actions, and I've just realised there's a bit of a migration problem here... the way I've done it at the moment, players who currently have god-hated equipment on them will just be able to use it for the rest of their game after they upgrade, without punishment. Some options, none of which I love: * Force-removing the forbidden equipment on upgrade - 23:00:44 quite unkind and could have some extremely nasty edge cases. * Docking piety over time for forbidden equipment, after a warning on upgrade - a bit less unkind but still could be quite tough on the player * Mitigating this by maintaining the conduct code for a while, so that gods are both forbidding and punishing things Better ideas would be very welcome 🙂 23:03:01 <04dracoomega> Frankly, in situations like this, I am content to let players just keep wearing them, if fixing this is actually difficult. It's only what they would have been allowed to do on that game one day earlier. Feels low-harm to me. 23:03:35 <11Odds> No it does actually make them more powerful - e.g. the rampagers of chei are just free to rampage 23:04:00 <04dracoomega> (Let them, imo. Someone will probably have fun with it.) 23:04:36 <11Odds> Sounds good to me 🙂 23:04:43 <04dracoomega> Like, I don't mean to sound flippant, but these 'transitional' games are very small in number and I think 'harmless' consequences are not worth significant code effort to avoid them 23:04:50 <02Darby> benefits of trunk 23:05:22 <11Odds> Yeah that makes a lot of sense 23:05:24 <04dracoomega> The main goal of save compat is to make sure nothing breaks. That the player can keep playing without their game being lost or otherwise heavily impaired somehow. If doing less means players get a little more, generally I am fine with that. 23:05:43 <04dracoomega> It's 'exploitable' for like a day 23:06:23 <04dracoomega> (Nobody was ever prevented from using draining weapons with Ely for power reasons, anyway) 23:06:50 <11Odds> Right, one game per player can rampage with chei is hardly a massive thing (and some other stuff, but that's the one I mostly expect to happen) 23:08:28 <11Odds> Cool thanks, I shall go back to testing all this... 23:10:21 <09hellmonk> thanks for taking on this project 23:10:39 <11Odds> Forbidding rather than penance is turning out much nicer codewise I think - we no longer have to keep warnings in sync with actual punishments in two completely separate places. 23:11:36 <11Odds> I've been slightly surprised not to find any existing bugs where penance warnings and penance don't match up 🙂 23:13:08 Unstable branch on cbro.berotato.org updated to: 0.35-a0-467-g0a0f4f6124 (34) 23:15:11 <04dracoomega> I mean, there are a couple cases where you can currently still get penance without a warning 23:15:16 <04dracoomega> I think? 23:15:57 <04dracoomega> (I hate to say that without saying what it was, but I feel like when I was last looking into doing the work you're doing now, I found at least one or two very niche cases where it was possible to get good god penance without a warning. I just don't remember what.) 23:16:15 <04dracoomega> Those get automatically fixed by making that not be a thing, of course 23:17:42 <02Darby> bestial takedown with a hated weapon was the most recent one? 23:18:41 <04dracoomega> You may be correct 23:18:52 <04dracoomega> Sounds familiar 23:22:52 <11Odds> Yeah definitely this work wouldn't systematically spot all such cases 23:23:26 <11Odds> But I kind of expected to stumble across an evil unrand that wasn't actually giving penance, or something 23:56:00 Monster database of master branch on crawl.develz.org updated to: 0.35-a0-467-g0a0f4f6124