02:15:36 Unstable branch on cbro.berotato.org updated to: 0.27-a0-54-g64bdb92e38 (34) 03:03:06 -!- amalloy is now known as amalloy_ 03:10:55 Stable (0.26) branch on cbro.berotato.org updated to: 0.26-b1-52-g61e05abef2 03:31:11 Stable (0.26) branch on crawl.kelbi.org updated to: 0.26-b1-52-g61e05abef2 03:31:25 Fork (bcrawl) on crawl.kelbi.org updated to: 0.23-a0-3840-g9efd504fd6 03:32:11 Fork (bcadrencrawl) on crawl.kelbi.org updated to: 0.03-86-g9e5655dcad 08:26:28 ZureaL: looks like console spectating (but not playing) is broken on cbr2. If you have time to take a look at that one, would appreciate it 08:26:37 when trying to spectate it just drops the connection 08:26:51 huh 08:26:53 iirc there's a common config issue that tends to cause this, but it's hard to figure out 08:27:04 like bad perms on a socket? or something 08:27:15 so it's not a webtiles problem, just a console problem? 08:27:23 yep, specifically watching users from console 08:27:30 that's managed by dgl, not webtiles 08:27:35 and playing works ok 08:27:44 all versions or just 0.26? 08:27:54 well watch isn't version-specific in any way 08:28:09 it's just like this thing that starts copying ttyrec output to the watchers terminal 08:28:16 it doesn't interface with crawl directly in any way 08:28:20 sorry, I'm going to blame advil for all this, he set up the console access :). Let me see if I can poke around 08:28:36 haha, yeah maybe advil can recall a fix, I think he's maybe seen it before too 08:28:49 I should check my cdi server actually 08:28:59 I know playing works but not sure if spectating does, maybe I can help diagnose 08:32:18 I'm seeing a "cannot open inprogress-dir" in the dgldebug.log 08:36:35 ah 08:36:38 hrm 08:39:47 sorry, had to feed the dog...I'm pretty sure it's something I did in the dgamelaunch.conf file when I implemented 0.26, I just gotta find it 08:40:02 ZureaL: perms on that dir? on my server I have ownership crawl:crawl, perms 755 on the dgldir/inprogress dir as well as all the version-specific subdirs 08:41:24 yea, owned by crawl:crawl and 755 08:41:24 I notice that the intructions have you make them using sudo, which actually wouldn't give them those perms; my console spectating does work 08:41:24 hmm 08:41:24 |amethyst: advil: any idea why dgl would be unable to allow spectators if console playing is working; dgldebug.log warns about opening the inprogress-dir 08:41:28 all the subdirs there for the various versions are also crawl:crawl 755 08:42:05 oh I guess one question is how that dir is configured in dgl conf, but I think those are specified per game config, not server-wide 08:42:14 right 08:43:02 for example; inprogressdir = "%rinprogress/crawl-git/" 08:43:54 yep, same as I have 08:45:01 ZureaL: well, I wish I could remember what the fix was last time this came up. I'll try to figure that out and ping you again (and maybe add this to the setup doc so we don't forget next time) 08:45:14 right on, I'll keep poking 08:53:34 okay, I was missing an inprogress dir for crawl-26-seeded, so I added it and made sure ownership perms are right 08:53:40 care to try again? 08:59:31 ZureaL: same thing, it just drops the connection 08:59:41 ZureaL: do you still get that dgl error? 08:59:52 I'm guessing that error and the problem may not be related 09:00:08 grrr 09:00:27 I actually don't have that subdir either (guess I should create it) 09:00:28 yea, still getting the cannot open inprogress-dir 09:00:41 I had that one for 24 and 25, just not 26 09:01:28 yep, I have it for 25 and git, not 26 09:01:51 let me see if the crawl-dev logs have anything 09:02:52 this just proves that webtiles is superior to console 09:06:40 FALSE! 09:06:55 there's something about an icps command to fix things, I'm looking into that 09:08:02 ok 09:09:01 ZureaL: https://dpaste.com/29LJUSNJD 09:09:10 this fix from |amethyst 09:10:02 that's a weird friggin fix 09:10:02 aye; I don't actually get anything listed for user crawl on my system 09:10:02 do you? 09:10:02 with the icps -m command? 09:10:02 that part just lists it 09:10:04 standby... 09:10:12 on a conf call and I'm presenting 09:10:16 haha, sorry 09:10:33 Tell them I have some amazing new data about vault kills 09:11:14 I'm sure they'll look forward to diggin into it 09:14:43 I do get a result with icps -m 09:16:28 daily scrum meetings and I'm the product owner, so those can get...interesting at times 09:17:21 ah, if you do then perhaps the icprm -M command is a good idea 09:27:25 so I'm looking at this icps stuff and I'm not sure where the 2nd key came from to use with the -M part of the command 09:29:18 oh, nm, I'm reading it right now 09:29:44 I think there's only one key? 09:29:57 neil's example only shows one, as the first field 09:30:18 okay, try now? 09:30:45 alas, same thing 09:30:52 just disconnects 09:36:00 johnstein said this fix didn't always work for him either 09:36:05 but apparently it did at times 09:36:06 I could try rebooting the box 09:36:30 I seem to recall having an issue initially on CDI but I can't remember what I did to fix (and am not 100% I did have this bug) 09:37:26 certainly possible that this works, although perhaps worth attempting at a less busy time? I wish I had a better way to diagnose this 09:37:52 I'd like to get it fixed before the tourney starts tomorrow 09:38:45 Sure, I would say this affects a pretty vanishingly small number of users (to be clear, console users can still play ok), but perhaps worth a shot 09:39:05 I see that when I spectate myself on cdi, I do finally get a shared memory key created 09:39:16 we'll see if advil ever shows up 09:40:16 interesting, removing it using icprm -M didn't disrupt my watching 09:40:33 ZureaL: out of curiosity, do you still have an icps -m entry? 09:40:48 I do not 09:40:54 looks like the key deletion did work in some way, as the key listed has changed for me 09:41:07 yes, mine has changed whenever I used that command 09:41:11 just didn't fix the problem 09:41:36 rip 09:41:50 new key just got installed 09:42:40 nice, there is a page for diagnosing stuff: https://crawl.develz.org/wiki/doku.php?id=dcss:server:maintaining_dgamelaunch_and_webtiles 09:43:02 sadly only mentions the fix you already did 09:47:39 can you try again? 09:48:02 I notice that in the deployment of a new stable release, dglwhere needs restarted and I hadn't done that 09:48:33 -!- Tuxedo[Qyou] is now known as BaitBot 09:50:13 ZureaL: alas, same disconnect 09:50:13 shitballs 09:50:13 notice there's a good deal of delay when connecting with user crawler 09:50:38 but starting up a game does work 09:51:03 it's very responsive after getting to the dgamelaunch terminal, but initial authentication is pretty slow 09:51:43 makes me wonder if there's a disconnect between that crawler user and the crawl user (I assume you use a user id of 'crawl' and not 'crawler' for DGL) 09:52:07 well, cao uses "joshua" and that one works, at least 09:54:50 I also failed to update the trigger-rebuild script, but I dont think that's called outside of the admin menu 09:54:54 but that's fixed now 09:55:57 there's a crawler user for console 09:55:57 crawl, crawl-dev and crawler 09:57:29 yep, and I think crawler just runs dgamelaunch and that gets setuided 09:57:47 so probably not an issue with that login per se 09:58:08 ZureaL: any issues you're seeing from the sshd log, by any chance? 09:59:38 Jan 7 08:24:47 nethack systemd[1]: Started Session 2679 of user crawler. 09:59:38 Jan 7 08:24:50 nethack systemd[1]: session-2679.scope: Succeeded. 09:59:42 just stuff like that 10:00:49 nethack systemd...traitor! 10:01:41 yeah I really don't know, maybe advil or neil has a clue. I will say that this is a pretty minor issue, as console users can still play and they can still spectate via webtiles, at least 10:01:41 true story, this also runs nethack 10:01:56 imagine the plot twist where that turns out to by why spectating doesn't work 10:02:11 s/by why/be why 10:13:09 I started out running a nethack server before johnstein and I talked about moving CBRO. I'm running a couple NH versions but no one uses them except for me and a couple of friends 10:29:12 I've been told that spectating has never worked for cbr2 so it might not be a recent thing 10:41:50 I'm not sure if anyone is planning on doing release today, but I wouldn't have time to do myself until tomorrow morning 10:42:01 which is fine with me, but if anyone strongly preferred a release today instead of tomorrow, you might want to do it yourself 10:42:07 well I found *a* problem, some of the inprogress directories were missing a - 10:42:13 didn't fix it though 10:42:25 (in dgamelaunch.conf) 10:42:30 hrmmm 10:43:22 I don't have any of those inprogress dir errors in my dgldebug.log fwiw 10:44:03 I do have: purge_stale_locks dir open failed 10:46:00 I just got that same error in CBR2's logs 10:46:14 that was what happened when I tried to start a seeded game 10:47:34 (it was the seeded inprogress dirs that were wrong) 10:47:56 the dgl update-stable subscript helpfully makes the dirs for zotdef but not for seeded 10:48:21 I just need to push some fixes for that to the repo, I guess 10:49:12 also, for some reason the 4 submenu isn't working for me, not obvious why 10:49:44 I pushed the 0.26 version on CBR2 by myself, so it's possible I fucked something up 10:49:58 ok, that menu does work for non-admin users 10:50:32 ah wrong dgl command 10:53:19 aha the seeded inprogress dirs were still wrong, used the cao format 10:53:24 looks like watch is working now 10:56:07 nice! 10:56:35 did the dir names just not match the format the dgamelaunch.conf was expecting? 10:57:36 Is this connected to the way ttyrecs that haven't been compressed yet are 404s? (eg https://cbro.berotato.org/ttyrec/hatsuma/ ) 10:57:38 no, the dir names in dgamelaunch.conf were wrong and didn't exist 10:57:54 no, that's unrelated 10:59:03 gah, I should have mentioned checking the match between those to Zureal 10:59:03 well, at least now we know that a mismatch causes this specific behaviour 10:59:11 okay, so upstream variable for inprogessdir was wrong? 10:59:27 valuable in the inprogressdir variable...blah blah 10:59:39 value...fuck I cannot type this morning 11:05:16 gammafunk: see? I was right to blame advil. :) 11:05:59 blame advil at your own risk; if you blame him too much, he doesn't help you fix the unsolveable bugs 11:06:28 you blame me too much, the worst I can do is try to kill you with a vault when you play crawl 11:07:10 lol 12:06:34 Unstable branch on crawl.akrasiac.org updated to: 0.27-a0-54-g64bdb92 (34) 12:26:55 03gammafunk02 07* 0.27-a0-55-ge8fdfcf: Remove references to Charms magic (split) 10(29 minutes ago, 4 files, 19+ 19-) 13https://github.com/crawl/crawl/commit/e8fdfcfcf69b 12:27:49 03gammafunk02 07[stone_soup-0.26] * 0.26-b1-53-gfce45f2: Remove references to Charms magic (split) 10(30 minutes ago, 4 files, 19+ 19-) 13https://github.com/crawl/crawl/commit/fce45f26fad5 12:32:12 Unstable branch on crawl.kelbi.org updated to: 0.27-a0-55-ge8fdfcfcf6 (34) 13:21:12 <12e​bering> @PleasingFungus did you mean for fenstriders to have the melee-range agony? 14:13:42 finally, player-monster symmetry 14:17:39 <10P​leasingFungus> yes 14:18:04 <10P​leasingFungus> i thought it was an interesting threat 14:19:02 <10P​leasingFungus> very similar to para really - if you can keep your distance, you can survive having bad Will 14:23:20 <12e​bering> cool 15:18:02 considering rewording "Max HP Drain" (in ability descs) to just "HP Drain", any thoughts? 15:32:25 <10P​leasingFungus> unsure 15:32:44 <10P​leasingFungus> i like the shorter phrasing, but not sure it’d be clear to players 15:34:06 -!- aidanh_ is now known as aidanh 15:49:58 right now I think it reads as permanent 15:50:37 which maybe the shorter phrasing is further from implying? 16:02:04 I don't suppose anyone can identify the vault in this thread? (tw, picture of swastika vault): https://www.reddit.com/r/dcss/comments/ks7ztt/i_always_knew_there_was_something_off_about_this/ 16:45:23 <10P​leasingFungus> @advil i grepped around and couldn’t find anything obviously similar - assume there’s some substs or shuffling happening 17:00:15 03advil02 07* 0.27-a0-56-ge4986e8: Missed hailstorm animations from 64bdb92e382d 10(14 minutes ago, 1 file, 2+ 1-) 13https://github.com/crawl/crawl/commit/e4986e83c084 17:00:15 03advil02 07* 0.27-a0-57-g216b618: Prioritize sandblast over throwing stones for EE start (ardl) 10(13 minutes ago, 1 file, 21+ 0-) 13https://github.com/crawl/crawl/commit/216b6189a382 17:00:15 03advil02 07* 0.27-a0-58-g0ec5f8a: Fix some oddnesses with fire order from invalid quivers 10(2 minutes ago, 1 file, 3+ 3-) 13https://github.com/crawl/crawl/commit/0ec5f8ab9351 17:01:11 re that vault, it might be an old version too, or fork 17:01:21 we have removed at least one thing like that since I've been around 17:02:43 Unstable branch on crawl.kelbi.org updated to: 0.27-a0-58-g0ec5f8ab93 (34) 17:16:10 03advil02 07[stone_soup-0.26] * 0.26-b1-54-g40af067: Missed hailstorm animations from 64bdb92e382d 10(30 minutes ago, 1 file, 2+ 1-) 13https://github.com/crawl/crawl/commit/40af067e96c8 17:16:10 03advil02 07[stone_soup-0.26] * 0.26-b1-55-gf96bd0e: Prioritize sandblast over throwing stones for EE start (ardl) 10(29 minutes ago, 1 file, 21+ 0-) 13https://github.com/crawl/crawl/commit/f96bd0e9b059 17:16:10 03advil02 07[stone_soup-0.26] * 0.26-b1-56-g48207bc: Fix some oddnesses with fire order from invalid quivers 10(18 minutes ago, 1 file, 3+ 3-) 13https://github.com/crawl/crawl/commit/48207bc6663a 17:16:10 03advil02 07[stone_soup-0.26] * 0.26-b1-57-gb162d06: Improve coverage of disabling beam animations (12356) 10(24 hours ago, 2 files, 40+ 23-) 13https://github.com/crawl/crawl/commit/b162d0664afa 17:20:08 -!- jfcaron_ is now known as jfcaron 18:24:00 Unstable branch on underhound.eu updated to: 0.27-a0-58-g0ec5f8ab93 (34) 20:44:27 hm did I manage to get everyone off tornado 2.4? 20:44:36 iirc I didn't 20:52:26 Stable (0.26) branch on underhound.eu updated to: 0.26-b1-57-gb162d0664a 21:16:33 03advil02 07* 0.27-a0-59-gfbafd83: Attempt to handle WebSocketClosedError in Tornado 6 10(2 minutes ago, 1 file, 35+ 10-) 13https://github.com/crawl/crawl/commit/fbafd837a320 21:30:31 Unstable branch on cbro.berotato.org updated to: 0.27-a0-59-gfbafd837a3 (34) 21:30:38 %cache 21:30:39 CCACHE status: Hit Rate = 46.76 % | Files Cached = 3576 | Cache Size = 4.3 GB | Max Cache Size = 5.0 GB 21:32:22 Unstable branch on crawl.kelbi.org updated to: 0.27-a0-59-gfbafd837a3 (34) 21:43:44 looks like cwz has 0.26 21:43:56 haven't tested all the urls, but if they're good, we should be ready for t 21:44:08 I can wipe the test data and restart taildb 21:57:25 looks like they all work, have scripts in normal mode, but will do more stuff tomorrow morning before t 21:57:25 test data all gone 23:08:47 <10P​leasingFungus> oops, i failed to get wormcan's sprites in before cwz got 0.26 23:25:52 <09g​ammafunk> Rip 23:26:31 <09g​ammafunk> The good news is they all switch to trunk after t 23:30:21 <09g​ammafunk> %git 0.26.0 23:30:22 <04C​erebot> Could not find commit 0.26.0 (git returned 128) 23:30:54 <09g​ammafunk> !kw r 23:30:55 <04C​erebot> Built-in: r => race!= 23:30:59 <09g​ammafunk> Doh 23:31:11 <09g​ammafunk> !kw t 23:31:12 <04C​erebot> Built-in: t => start>='2021-01-08 20:00:00' time<'2021-01-24 20:00:00' ((cv=0.26|0.26-a)) explbr= file!=cwz/soup/trunk/milestones file!=cwz/soup/trunk/logfile