00:10:51 Stable (0.32) branch on cbro.berotato.org updated to: 0.32-b1-36-g01caadce97 00:34:52 <07acrobat> > If it's always going to innately have risks, I say it's much better to have a strong risky brand that at least is an option versus a weak risky brand that mostly serves as a trap. I Agree. > Brand weapon scrolls already present risks because they can provide one with a brand a lot weaker than one had on it already, like venom on earlygame two-handers or heavy on shortblades or draining on lategame characters or spectral on frailer 00:34:52 characters who don't want to take all the damage-sharing.) This sounds a bit weird to me. In the early game, all brands except chaos are better than no brands, so using it to your unbraded primary weapon has no risk at all (if there is no chaos). And if your weapon already have decent brand then you really don't have to rebrand it taking those risk. There's not much to be gained from such a gamble. (except for a few bottom brands, there's not much 00:34:53 difference in performance between brands imo. the best outcome is spectral for most char, which isn't super good like before after mutiple nerf imo.) > I personally would prefer there to be some chance of chaos being of interest to non-Xom characters Personally I am not a fan of chaos brand coming into non-xome game. Aside from it's risk, I think its inconvenience is not small problem. I have to be very careful when using it and even with RC, it causes 00:34:53 too much -more- (inv removal was a big improvement, but I still doubt it's tolerable.) 01:05:46 <04dracoomega> Your last statement makes me wonder if there was a slight misunderstanding. I don't think she was talking about adding more chance of chaos to things, but rather that the existence of the possibility of branding a weapon with chaos (which has virtually always existed, and even at the present rates for quite a number of years now) is better for the game than the brand virtually not existing outside of Xom. Compared to how things 01:05:46 have been for multiple years, new chaos is almost strictly better for the player than it used to be, while also being no more common. (From my own perspective, I am not sure I actually agree that chaos is worse than unbranded, even early on. Like, yes, it introduces the possibility of doing something bad to the player that didn't exist when the weapon was unbranded, but the average effect is considerably more likely to be helpful than harmful and is 01:05:50 sometimes extremely helpful. When a player loses due to a chaos effect, it's extremely obvious that this is what killed them, but that sort of glosses over numerous situations where the brand may have made the difference between a fight going extremely poorly and it going well. ie: "Just because it killed you doesn't mean that you wouldn't have died earlier if you didn't have it.") This may be somewhat less true for an extremely high winrate player 01:05:50 attempting long streaks, where you feel confident that you can handle anything the game throws at you if no additional variance is introduced, but even if we accept this to be statistically true, that applies to a very, very thin slice of the playerbase. 01:07:38 <04dracoomega> I hope this doesn't sound like brushing off your concerns, and I can appreciate if you don't enjoy using chaos weapons yourself, but I do think they have been solidly popular since the recent rework of them - and even among some players who are very good at the game also. 01:39:25 <07acrobat> Oh, what I meant was.... before 0.28, chance of chaos from brand scroll was much lower(1~2% iirc) and then it was increased a lot when brand scroll was simplified. This made chaos appear much more often in non-xom games, which I don't prefer. Chaos is worse than unbranded <- Actually I'm not very confident in this currently. Like I said, I don't have real experience using 0.32 chaos brand. My logic is just that players can handle 01:39:25 pretty much everything the game throws at them, so chaos brand that creates unnecessary variables is worse than no brand, like you already said. (Is this only true for a very small number of skilled players? I don't know.) If many people enjoy using chaos brand in non-xom games, I'm happy if my opinion is ignored 🙂 01:41:14 <06regret-⸸nde※> %git 75e6d9b 01:41:14 <04Cerebot> kate- * 0.28-a0-732-g75e6d9beac: Simplify scrolls of brand weapon (2 years, 10 months ago, 2 files, 26+ 21-) https://github.com/crawl/crawl/commit/75e6d9beac14 01:43:59 <06regret-⸸nde※> (...I do think it's weirdly communicated for "rarely" to mean "50% as much as 'commonly' for melee weapons and 33% as much as 'commonly' for ranged weapons".) 01:44:30 <06regret-⸸nde※> (Maybe "rarely" only for ranged's slice and "uncommonly" for melee?) 01:45:32 <04dracoomega> (I actually didn't even realize this was documented in any fashion in game, and kind of thought draining was on the 'common' list) 05:08:42 Unstable branch on crawl.akrasiac.org updated to: 0.33-a0-35-g516f9d4 (34) 06:41:24 <08nicolae> q: thinking about it later, wouldn't this have the same troubles as before, if it's on V:4 and/or hard? 8 - (d*2) is still nonpositive if d >= 4 06:43:39 <08nicolae> ...what's the order of operator precedence in lua 07:18:29 <09gammafunk> https://www.lua.org/pil/3.5.html 07:19:38 <08nicolae> okay, yeah, so the 8 - d * 2 will be taken as 8 - (d * 2) and not (8 - d) * 2. @regret-⸸nde※ which one did you have in mind? 08:22:31 Never gonna give you up... 08:28:22 I'm back 09:57:37 <04melvinkitnick> Hey, I have some weird behavior on scoring with a CXC user, if someone here have a suggestion. What they claims: > Endy73: @melvinkitnick Hi, there. I have issue with my latest win scoring: it didn't registered on the leaderboard and didn't displayed in my profile. The player name Endy73, character DEEE, game started on Aug 27th, Thank you in advance and also thank you for keeping to maintain our endless joy and suffering) 09:57:38 Their scoring page on CAO: http://crawl.akrasiac.org/scoring/players/endy73.html. See also attached screenshot of our recent conversation. Here is what I found in logfiles (here for 0.32) for a DEEE found in 2024: v=0.32-b1:vlong=0.32-b1-8-g0b4fe110a3:lv=0.1:vsavrv=Git::0.32-b1-8-g0b4fe110a3:vsav=34.282:tiles=1:name=Endy73:race=Deep Elf:cls=Earth 09:57:38 Elementalist:char=DEEE:xl=26:sk=Spellcasting:sklev=27:title=Archmage:place=D::$:br=D:lvl=0:absdepth=1:hp=174:mhp=174:mmhp=174:mp=41:mmp=50:bmmp=50:str=13:int=42:dex=16:ac=16:ev=29:sh=18:god=Vehumet:start=20240727125903S:dur=27149:turn=99523:aut=1000118:urune=4:nrune=4:kills=2300:maxskills=Spellcasting:fifteenskills=Fighting,Dodging,Spellcasting,Ice Magic,Earth 09:57:39 Magic:gold=5326:goldfound=8577:goldspent=3251:scrollsused=69:potionsused=39:sc=1880199:ktyp=winning:dam=-9999:sdam=0:tdam=0:piety=199:end=20240728122829S:map=dpeg_arrival_three_bigger_balls:seed=5619358695684568147:tmsg=escaped with the Orb:vmsg=escaped with the Orb and 4 runes! 09:57:39 <04melvinkitnick> https://cdn.discordapp.com/attachments/747522859361894521/1278398094488240238/Screenshot_20240828_182452.png?ex=66d0a881&is=66cf5701&hm=b916bc3a1cae29edc29ec275c3b87e4110ae633fb724a33c401b6923af15ee80& 09:57:41 <04melvinkitnick> It seems to be the aforementioned game, as everything else seems to be correct (also compared to their dungeon exit screenshot, scoring is the same, etc...). Morgue exists on server too: https://crawl.xtahua.com/crawl/morgue/Endy73/morgue-Endy73-20240828-122829.txt Now this is in 0.32 logfile, which didn't exist before I built this version on 24 Aug 2024, and this game was apparently started on 27 Jul 2024. So possibly a TZ issue 09:57:41 somewhere, though it's definitely weird. Though, look at this game I won on 17 Aug 2024: https://crawl.xtahua.com/crawl/morgue/melvinkitnick/morgue-melvinkitnick-20240817-164740.txt Although timestamps are also wrong, this game correctly appear on my scoring page. 09:57:42 v=0.32-a0:vlong=0.32-a0-2121-g385508daed:lv=0.1:vsavrv=Git::0.32-a0-2120-g8f138eace0:vsav=34.282:tiles=1:name=melvinkitnick:race=Octopode:cls=Shapeshifter:char=OpSh:xl=26:sk=Shapeshifting:sklev=25:title=Skinwalker:place=D::$:br=D:lvl=0:absdepth=1:hp=189:mhp=189:mmhp=199:mp=33:mmp=35:bmmp=35:str=38:int=14:dex=12:ac=17:ev=47:sh=0:god=Ru:start=20240715150820S:dur=17372:turn=63158:aut=613149:urune=5:nrune=5:kills=2568:fifteenskills=Fighting,Dodging,Unarmed 09:57:42 Combat,Translocations,Shapeshifting:status=storm-form,flying,hasted,studying 2 skills,lightly drained:gold=4032:goldfound=8570:goldspent=4538:scrollsused=58:potionsused=33:sc=3427703:ktyp=winning:dam=-9999:sdam=0:tdam=0:piety=160:end=20240717164740S:map=dpeg_arrival_water_temple_mockup:seed=13370981317706221815:tmsg=escaped with the Orb:vmsg=escaped with the Orb and 5 runes! 10:06:13 Stable (0.32) branch on crawl.akrasiac.org updated to: 0.32-b1-36-g01caadce97 10:09:39 <04melvinkitnick> It seems to be the case for all 0.32 games, see this list for instance. None of them games are on CAO. 10:09:39 <04melvinkitnick> https://cdn.discordapp.com/attachments/747522859361894521/1278401119315558471/message.txt?ex=66d0ab52&is=66cf59d2&hm=66f841adafa8df3b400a997a4c4918f18c71b8750dc021c86c4130bf096d0819& 10:15:26 <04melvinkitnick> Oh ! Perhaps 0.32 games are not scraped yet by CAO scripts ? That would be the simplest explanation 😄 10:55:42 <09gammafunk> Heh, sorry it's that simple! We haven't added 0.32 sources to scoring and will take cao scoring down during tournament to save resources 10:56:14 <09gammafunk> But 0.32 will get added afterwards and all previous 0.32 games will count 10:56:37 <04melvinkitnick> Great. Thanks 🙂 11:38:03 <06regret-⸸nde※> The point of the offered block is that it guarantees polterguardians by V:4. You can just raise the initial weights of the yaktaurs if you still want some of those earlier than that. 11:39:40 <06regret-⸸nde※> (...Actually, it still offers yaktaurs in the block.) 11:43:31 <06regret-⸸nde※> (I've done this sort of scaling with several other vaults already- even in that function, it drops white ugly things in the icebox set on V:4 and harpies in the garden set on V:4.) 12:05:35 Webtiles server stopped. 12:05:43 Webtiles server stopped. 12:05:45 Webtiles server stopped. 12:06:01 Webtiles server started. 12:08:35 Webtiles server stopped. 12:08:37 Webtiles server stopped. 12:08:44 Webtiles server stopped. 12:08:52 Webtiles server stopped. 12:08:59 Webtiles server stopped. 12:09:09 Webtiles server stopped. 12:09:13 Webtiles server started. 12:41:31 <08nicolae> hmm! i guess i would need to tweak my vault a little if the ranges are set up so that at v:4 (or v:3 hard) all the 1 glyphs are supposed to be polterguardians 12:43:13 <06regret-⸸nde※> The offered block gives 1/3 yaktaurs and 2/3rds polterguardians on V:4. 12:46:33 <08nicolae> ah, you are correct 12:46:48 <08nicolae> i misremembered the yaktaur formula (yaktormula) 13:27:08 <04dracoomega> Looking through github issues again, I've tried reproducing https://github.com/crawl/crawl/issues/3995 myself, and neither a WSL build nor a windows local tiles build gets the crash they're talking about. I can't really begin to guess where in the relevant function the crash could be happening without more debug data or the ability to reproduce it myself. Given that it's in UI layout code, I'm a little worried it could be even more 13:27:09 narrowly platform-specific than that. I don't suppose anyone here is running Arch Linux and could give the save file a try in a local tiles build there? 13:27:39 <04dracoomega> To at least see if it's a general problem or more specific to the person who posted the issue 13:48:53 Leissi (L20 FoFi) Crash caused by signal #6: Aborted (D (Sprint)) 13:50:58 <12ge0ff> yeah, i can confirm: starting a fresh GnSh and opening the Mutations - Form Properties screen crashes the game (it happens only on Local Tiles; Web Tiles and console builds don't crash) 13:51:18 <12ge0ff> (that's in the latest trunk) 13:52:15 <04dracoomega> And this is on Arch Linux as well? 13:52:34 <12ge0ff> kinda 🙂 (on an Arch-derivative) 13:52:54 <04dracoomega> Good to know. Well, bad to know, really. How the heck do I even start to work on this one? ^^; 13:53:16 <04dracoomega> (Thank you, though) 13:57:04 <04dracoomega> Nothing in the crashing function itself seems to have changed in 3 years. I don't immediately see how my future-mutation code on another tab there could be influencing it 14:19:35 <12ge0ff> "good" news: the crash was introduced earlier: i tried 0.32-a0-1927-ga117348762, the version before your "Display future XL-based mutations on the 'A' screen (elliptic)" commit, and it crashes too! (for GnSh and for HuSh) 14:21:54 <12ge0ff> Were there other 0.32 mutation screen-related changes? 14:22:29 <04dracoomega> Not that I'm immediately aware of, but I could be wrong 14:25:59 <12ge0ff> %git 44e7262d6 14:25:59 <04Cerebot> NormalPerson7 {DracoOmega} * 0.32-a0-1110-g44e7262d61: Improve the A screen for forms (5 months ago, 6 files, 220+ 70-) https://github.com/crawl/crawl/commit/44e7262d61be 14:26:41 <12ge0ff> ^-- could be this commit (currently building 0.31 to check if it crashed too) 14:27:26 <04dracoomega> (I am currently going through the long process of seeing if I can get Arch working over here) 14:27:41 <04dracoomega> Not that I can be at all confident if doing so through WSL will even have the same crash 14:28:04 <12ge0ff> oh, you can use different distros with WSL? 14:29:30 <04dracoomega> To an extent (I think this was a 3rd-party project?) 14:30:38 <04dracoomega> I mean, my own familiarity with WSL is admittedly mostly limited to "I managed to find enough instructions to follow to eventually get Crawl's build environment to work for me" ^^; 14:31:09 <04dracoomega> The amount of 'how to operate linux' I know outside of that is remarkably sparse 14:39:00 <12ge0ff> Welp, 0.31.0-35-g6cd5f5668d crashes too on opening Form Properties. It's surprising no one reported it earlier 14:41:42 <04dracoomega> Perhaps the number of people playing local tiles on this distro family in particular (and also realizing that a particular UI sub-menu that is never actually needed to do anything exists) is very small 14:43:41 <04dracoomega> Oh, that predates talismans, even, then, doesn't it 14:43:54 <04dracoomega> Nevermind, that's stable, isn't 14:43:59 <04dracoomega> Not a0-35 >.> 14:44:39 <04dracoomega> (I actually don't know if that menu was only added in any form after talismans or if spell forms used it in some fashion) 14:45:18 <12ge0ff> %git cb8bc90ed4a8 14:45:19 <04Cerebot> PleasingFungus * 0.31-a0-955-gcb8bc90ed4: Describe forms in the A-menu (10 months ago, 4 files, 266+ 98-) https://github.com/crawl/crawl/commit/cb8bc90ed4a8 14:45:26 <04dracoomega> (I'd actually forgotten it existed at all until this bug report >.>) 14:45:51 -!- TAS-2012v is now known as TAS_2012v 15:04:21 <12ge0ff> Form Properties screen has been crashing since it was added in this commit 15:05:48 <09gammafunk> damgot 15:06:12 <09gammafunk> DO's yred rework broke tournament scripts! > crawl@crash:~/dcss_tourney> Error processing: vlong=0.32-b1-36-g01caadce97:ac=25:gold=233:int=12:nrune=0:tiles=1:absdepth=13:char=DsAr:vsav=34.282:ev=15:sk=Maces & Flails:xl=15:title=Bludgeoner:god=Yredelemnul:lv=0.1:start=20240828143543:race=Demonspawn:aut=221329:dur=11635:type=uniq:mhp=123:cls=Artificer:status=under a protective aura,wreathed by umbra,studying Air 15:06:13 Magic:raceabbr=Ds:fifteenskills=Maces & Flails:mmhp=123:kills=919:src=cbr2:urune=0:hp=25:v=0.32-b1:mmp=24:sklev=15:verb=uniq:br=D:milestone=bound Azrael.:goldfound=2680:dex=16:bmmp=24:noun=Azrael:name=Midpoint:vsavrv=Git::0.32-b1-36-g01caadce97:potionsused=8:turn=21798:sh=11:place=D::13:mp=16:str=20:time=20240828210550:lvl=13:scrollsused=36:goldspent=2447 > Traceback (most recent call last): > File "taildb.py", line 60, in > 15:06:14 tail_logfiles(loaddb.LOGS, loaddb.MILESTONES, crawl_utils.UPDATE_INTERVAL) > File "taildb.py", line 28, in tail_logfiles > interval_work(cursor, interval, master) > File "taildb.py", line 13, in interval_work > master.tail_all(cursor) > File "/home/services/crawl/dcss_tourney/loaddb.py", line 388, in tail_all > oldest.process(cursor) > File "/home/services/crawl/dcss_tourney/loaddb.py", line 241, in process > 15:06:14 self.processor(cursor, self.filename, self.offset, self.xdict) > File "/home/services/crawl/dcss_tourney/loaddb.py", line 1227, in add_milestone_record > handler(c, d) > File "/home/services/crawl/dcss_tourney/loaddb.py", line 1126, in add_unique_milestone > extract_unique_name(game['milestone'])) > File "/home/services/crawl/dcss_tourney/loaddb.py", line 1093, in extract_unique_name > return 15:06:14 strip_unique_qualifier(R_KILL_UNIQUE.findall(kill_message)[0]) > IndexError: list index out of range 15:06:33 <09gammafunk> "bound Azrael" is unexpected milestone syntax 15:06:38 <04dracoomega> Well, in over an hour of trying to get an arch build working for myself, I have made it to the 'Crawl finally claims to have built, but will not run' stage >.> 15:06:39 <09gammafunk> (should be an easy fix) 15:07:30 <09gammafunk> hrm, feels extremely unlike to be disto-specific, re: arch linux build 15:07:35 <04dracoomega> libEGL warning: MESA-LOADER: failed to open vgem: /usr/lib/dri/vgem_dri.so: cannot open shared object file: No such file or directory (search paths /usr/lib/dri, suffix _dri) (Google is thus-far being less helpful for this than I'd hope) 15:08:26 <09gammafunk> I can try the save in my ubuntu certainly 15:08:38 <04dracoomega> No save needed. It just crashes in a fresh game in local tiles, apparently 15:09:01 <12ge0ff> yeah, just start a shapeshifter and open the A-form properties screen 15:09:26 <04dracoomega> (Ubuntu is what I've been using via WSL, but of course it's impossible for me to tell whether or not the 'through WSL' part is what causes it not to crash or the 'Ubuntu' part) 15:09:48 <09gammafunk> I see that this was commit 6 of 0.33 15:09:57 <09gammafunk> are we also sure that there wasn't somehow a subsequent fix? 15:10:01 <09gammafunk> that would render it unable to crash 15:10:19 it was reported on trunk, no? 15:10:27 <04dracoomega> It was reported on trunk, yes 15:10:34 <12ge0ff> is there gonna be more info in the crash log if i use a debug/debug-lite build? 15:10:43 <09gammafunk> I just mean that we're now at 15:10:45 <09gammafunk> %git 15:10:46 <04Cerebot> DracoOmega * 0.33-a0-35-g516f9d4634: Fix enhanced breath damage not displaying for red draconians (Undo) (21 hours ago, 1 file, 3+ 1-) https://github.com/crawl/crawl/commit/516f9d463415 15:10:52 <09gammafunk> some 29 commits later 15:11:39 <09gammafunk> another thing you can try: downloading the windows tiles build for that from cdo if...oh wait, that probably broke upon the version bump to 0.33-a 15:11:44 <09gammafunk> we always forget to do that initially 15:11:47 <04dracoomega> Crash still occurs after rebuilding for 0.33-a0-21-g2325b17bb0. (But I mean, I don't see what I would have done that would have addressed such a long-standing bug by accident) 15:12:10 <09gammafunk> ah I see subsequent comment 15:12:37 <09gammafunk> %git 9311ea8 15:12:38 <04Cerebot> DracoOmega * 0.33-a0-14-g9311ea8977: Fix a niche crash with displaying future XL-based mutations (4 days ago, 1 file, 1+ 1-) https://github.com/crawl/crawl/commit/9311ea8977ee 15:12:59 <04dracoomega> I mean, it may be possible to download something pre-built for Arch, but the point is that if I want to diagnose the problem, I want to be able to actually do stuff to the code in an environment where the crash occurs 15:13:13 <04dracoomega> (It does seem to have built, but won't currently run) 15:13:40 <09gammafunk> I think the distro is extremely unlikely a relevant factor; we have just seen vanishingly few distro-specific bugs 15:13:43 <09gammafunk> I honestly can't even recall a single one 15:14:10 <12ge0ff> it could be a linux-specific crash, not arch-specific 15:14:16 <09gammafunk> sure 15:14:23 <09gammafunk> but linux was already tried by DO 15:14:38 <09gammafunk> anyhow I'm building sdl tiles right now and will test 15:14:45 <09gammafunk> going to start with current trunk 15:14:47 fwiw it doesn't crash here (Ubuntu 22.04) 15:14:50 <09gammafunk> ... 15:15:27 <09gammafunk> In file included from fontwrapper-ft.cc:6: fontwrapper-ft.h:9:10: fatal error: ft2build.h: No such file or directory 9 | #include | ^~~~~~~~~~~~ compilation terminated. make: *** [Makefile:1806: fontwrapper-ft.o] Error 1 make: *** Waiting for unfinished jobs.... glwrapper-ogl.cc:25:13: fatal error: GL/glu.h: No such file or directory 25 | # include | ^~~~~~~~~~ have I 15:15:28 never set up tiles prereqs on this comp... 15:15:41 <09gammafunk> gonna need to look those up I guess 15:15:53 <09gammafunk> (anyone have the debs off the top of their head?) 15:16:13 <09gammafunk> ok I found 15:16:55 <09gammafunk> interesting, still get the GL/glu.h error 15:17:39 <09gammafunk> O 15:17:52 <09gammafunk> er, I'm on 24.04 and apparently I need something gl-related for tiles 15:18:35 sdl is built on top of opengl 15:23:35 <09gammafunk> (apparently 24.04 needs libglu1-mesa-dev) 15:24:57 <09gammafunk> no crash for me in latest trunk in 24.04, going to try that 6th commit 15:25:19 <09gammafunk> ...wait, sorry 15:29:56 I tried both HEAD and 2325b17bb0 15:29:59 <09gammafunk> @dracoomega tested latest trunk and that 6th commit the user was initially on in ubuntu 24.04, no crashes 15:30:37 <04dracoomega> Meaning it probably is somehow distro-specific 15:30:56 <09gammafunk> still don't think it is 15:31:14 <09gammafunk> there's a lot of things that can have gone wrong at the user level potentially 15:31:16 <04dracoomega> I mean, ge0FF said it crashes for them on Arch, no? 15:31:21 <09gammafunk> oh, really? 15:31:21 remember that Arch-style distros are rolling release, meaning they have a possibly much neweer SDL than Ubuntu will 15:31:23 <04dracoomega> On multiple versions 15:31:54 <09gammafunk> ah 15:31:56 <04dracoomega> (I mean, that is the entire reason I have been beating my head on trying to get this working on my end, after confirming it does crash there) 15:32:05 <09gammafunk> so it's possibly just a new sdl issue 15:32:25 <04dracoomega> You say 'just' as if that isn't slightly terrifying 15:32:53 <09gammafunk> it's moderately so! I bet any fix would be pretty simple. I'm wondering if it's easy to do a linux contrib build 15:33:12 <09gammafunk> I've not really tried, but that'd be one way to get a new SDL compared to the system one 15:33:40 <09gammafunk> but maybe just getting arch working is actually easier 15:35:24 wouldn't building sdl from contribs be an older version? submodules are pinned and the pin hasn't been updated in years 15:37:16 <09gammafunk> well, you could locally update the contrib to a newer checkout of the source (maybe adding upstream as a remove) and update the pin. But that's probably somewhat complicated because in the case of SDL there are possibly a bunch of other SDL contribs you'd also have to do that to 15:37:47 SDL2 SDL2_mixer SDL2_image 15:38:08 Unstable branch on underhound.eu updated to: 0.33-a0-35-g516f9d4634 (34) 15:42:32 <12ge0ff> I am using sdl from a system package (v2.30.6, the latest one), not the one from contribs 15:44:18 <09gammafunk> yeah, just talking about ways to potentially get DO a build of the game with your SDL 15:45:15 I have 2.0.20 fwiw (Ubuntu's perennially old) 15:45:15 <09gammafunk> no help to the current situation, but I should probably update INSTALL to mention that newly required libglu1-mesa-dev package for ubuntu, since it seems that wasn't required in past versions 15:46:35 <09gammafunk> I have 2.30.0 15:46:44 <09gammafunk> but I'm running the latest LTS 15:47:32 <04dracoomega> 40 minutes later and I have made absolutely 0 progress on getting Crawl to launch 15:48:20 <04dracoomega> It just prints libEGL warning: MESA-LOADER: failed to open vgem: /usr/lib/dri/vgem_dri.so: cannot open shared object file: No such file or directory (search paths /usr/lib/dri, suffix _dri) libEGL warning: MESA-LOADER: failed to open vgem: /usr/lib/dri/vgem_dri.so: cannot open shared object file: No such file or directory (search paths /usr/lib/dri, suffix _dri) libEGL warning: NEEDS EXTENSION: falling back to kms_swrast and then 15:48:21 hangs forever when I try. No similar problem I can search up has had a working solution. 15:48:50 <09gammafunk> not going to work on this now, but I try my "update SDL contrib" idea tomorrow and see if I can reproduce 15:49:21 <09gammafunk> feels like it might be a useful technique to have to diagnose these issues without resorting to installing distributions in vms and such 15:49:47 -!- TAS-2012v is now known as TAS_2012v 15:51:11 <09gammafunk> (although I seem to recall once trying to make a linux build using contribs once and it not going well due to our build system being what it is) 15:51:23 <09gammafunk> (once once) 15:54:29 <09gammafunk> also, just for the record, what are your versions of sdl2_mixer and sdl2_image? 15:55:58 <12ge0ff> sdl2_image: 2.8.2, sdl2_mixer: 2.8.0 15:56:03 <09gammafunk> ty 15:56:12 <09gammafunk> those are same versions as my system 16:41:18 <12ge0ff> So, if i comment out everything inside mutation.cc:update_form(), except for the first cpp add_entry(new MenuEntry("", MEL_ITEM, 1, 0)); // XXX spacing kludge? line, i get a crash 16:42:07 <12ge0ff> BUT, if i change this line to cpp add_entry(new MenuEntry("test", MEL_ITEM, 1, 0)); // XXX spacing kludge?, i get 16:42:08 <12ge0ff> https://cdn.discordapp.com/attachments/747522859361894521/1278499890212900873/image.png?ex=66d1074f&is=66cfb5cf&hm=eae423fda142c6e7eeb3fbdc91ab8b654daabf7a0caa95529f634855a190e48f& 16:45:15 <04dracoomega> Well, that's already a lot more than I've accomplished in my almost 3 hours of failing to get Crawl to work on Arch over here. (I really am close to just tearing my hair out altogether by this point, quite frankly >.>) 16:51:35 <12ge0ff> There are three lines with "// XXX spacing kludge?" in update_form(). Without them the form looks like this (and doesn't crash): 16:51:36 <12ge0ff> https://cdn.discordapp.com/attachments/747522859361894521/1278502274334851143/image.png?ex=66d10987&is=66cfb807&hm=5863d4b54d723001573742f164e1064039824b3026ec7d47c36fd74cf2c11754& 16:52:36 <12ge0ff> (the other way to not crash, is to change the first parameter of new MenuEntry from "" to "") 16:52:56 just out of curiosity, what hapopens if you use " " instead of ""? 16:53:06 <12ge0ff> with a space it still crashes 16:56:35 <12ge0ff> it crashes with two spaces two, but, with three spaces it works fine: add_entry(new MenuEntry(" ", MEL_ITEM, 1, 0)); // XXX spacing kludge? gives 16:56:35 <12ge0ff> https://cdn.discordapp.com/attachments/747522859361894521/1278503522521841720/image.png?ex=66d10ab1&is=66cfb931&hm=cbef59f2c951b0725f79c475939e2d25f2ef84c31b6f98a1e12436fd6e0389bf& 16:57:16 <12ge0ff> there's probably some width calculation that needs a long enough string 17:19:19 <12ge0ff> and, by the way, Vampires crash on the Blood properties screen too, w/o any forms 17:19:50 <04dracoomega> I... I just have to stop. I've spent like 7 hours in a row before peering into ancient Abyss bugs and that did less damage to my sanity than trying to get Crawl to launch on Arch over here has done. I'm not even exaggerating. 17:20:29 <04dracoomega> I feel like I have spent entire days worth of dev energy on it by now >.> 17:20:30 <12ge0ff> (and the "fix" is the same, set label to at least 3 chars/spaces in update_blood(): cpp for (int y = 0; y < lines; y++) // lines (properties) { string label = " "; // <--- here ) 17:21:33 <04dracoomega> (I tried looking at the menu layout code that might be breaking in response to the stuff you've discovered, but I'm just too spent T.T) 17:22:10 <09gammafunk> (Now's probably not a good time to ask DO if she's considered switching to GNU/Linux) 17:23:25 <09gammafunk> oh right, that milestone crash 17:25:52 <12ge0ff> it might be easier to setup a VM with Arch than to wrestle with WSL-Arch 17:32:37 <04dracoomega> At this rate: quite possibly! This was supposed to be the easier and quicker option. 17:32:44 <12ge0ff> also, which one are you using, https://github.com/yuk7/ArchWSL ? 17:33:33 <04dracoomega> Yes 17:38:33 <09gammafunk> in mon-death.cc:583 (record_monster_defeat()) cpp if (mons_is_or_was_unique(*mons) && !testbits(mons->flags, MF_SPECTRALISED)) is that spectralized condition to not trigger a second milestone if a lost soul is involved? 17:39:49 <09gammafunk> ah, I see that binding is its own, second milestone, good 17:40:10 <12ge0ff> Hmm, so if Form properties and Blood properties screens crash when adding a short/empty MenuEntry, this means all other MenuEntry("", ) calls could crash too 🙀 17:40:41 <09gammafunk> so this is easy, I can just do what the other milestone types do and massage it into a new milestone verb 17:41:39 <12ge0ff> (there are 8 more such calls, not counting those that have MenuEntry(something, where something can be empty) 17:50:56 <09gammafunk> we have a new mark milestone as well, don't we 17:51:52 <09gammafunk> !lm * file~~32 mark 17:51:53 <04Cerebot> No milestones for * (file~~32 mark). 17:52:01 <09gammafunk> !kw mark 17:52:02 <04Cerebot> Built-in: mark => verb=mark 17:52:13 <09gammafunk> !lg * file~~32 17:52:14 <04Cerebot> No games for * (file~~32). 17:52:18 <09gammafunk> oh right, not in sequell yet 17:53:34 <09gammafunk> I think that doesn't matter for tournament, we just record the milestone and do nothing with it 19:23:46 <09gammafunk> !lg * cdi 19:23:49 <04Cerebot> 134642. qw the Cudgeller (L4 OnGl), blasted by an orc priest (nerve-wracking pain) on D:3 on 2024-08-29 02:22:40, with 87 points after 2061 turns and 0:01:40. 20:24:10 <09gammafunk> !hs qw day 20:24:11 <04Cerebot> 271. qw the Conqueror (L25 OnGl of Okawaru), blasted by a golden dragon (bolt of cold) on Zot:1 (regret_index_zotdec_ceiling_fans) on 2024-08-28 12:30:18, with 505946 points after 59375 turns and 2:14:56. 23:45:15 Monster database of master branch on crawl.develz.org updated to: 0.33-a0-35-g516f9d4634