03:33:32 Experimental (bcrawl) branch on underhound.eu updated to: 0.23-a0-5208-geafff8c3b6 06:21:08 03dolorous02 07* 0.32-a0-1771-g4f39dfde82: Add another noisy hell message. 10(6 minutes ago, 1 file, 1+ 0-) 13https://github.com/crawl/crawl/commit/4f39dfde8275 10:07:45 New branch created: pull/3934 (1 commit) 13https://github.com/crawl/crawl/pull/3934 10:07:46 03Aliscans02 07https://github.com/crawl/crawl/pull/3934 * 0.32-a0-1772-g096704fd8e: Improve wrapping for monster_item_view_coordinates. 10(2 minutes ago, 1 file, 7+ 6-) 13https://github.com/crawl/crawl/commit/096704fd8e31 10:12:07 <12a​sciiphilia> python # CNC Server's config.py # how often to check for an active connection while playing; this has an effect on how often a connection is checked for basic life, as well as the following wo idle timer settings. connection_timeout = 600 # the maximum allowed idle time while playing. This timer is checked in intervals determined by `connection_timout`, so values will essentially be rounded up to that setting max_idle_time = 5 * 60 10:12:08 * 60 # Seconds until stale HTTP connections are closed # This corresponds to the tornado parameter `idle_connection_timeout`, which # will automatically close idle http connections that do not respond after a # period of time. Setting this to `None` gives the tornado default (1 hour). # This setting usually does not affect websockets connections, which use the # above webtiles-internal timeouts. http_connection_timeout = 600 kill_timeout = 10 # 10:12:08 Seconds until crawl is killed after HUP is sent It seems like the timeout settings are properly configured on the server, so why do I see such players in the lobby? 10:12:09 <12a​sciiphilia> https://cdn.discordapp.com/attachments/747522859361894521/1265355782174675075/image.png?ex=66a135e7&is=669fe467&hm=0f65cfe4e7dc3995c8bb18e5c52331d01fb7fd847e1521b9555c0f0b3bc8dbbb& 10:14:29 <12a​sciiphilia> Is there anyone who knows about this well? 10:40:59 03DracoOmega02 07* 0.32-a0-1772-ga91aa3b054: Don't let Xom door rings destroy items 10(17 minutes ago, 1 file, 1+ 1-) 13https://github.com/crawl/crawl/commit/a91aa3b0542c 10:40:59 03DracoOmega02 07* 0.32-a0-1773-g16c7f17c50: Don't be content with shifting items into closed doors 10(15 minutes ago, 1 file, 3+ 1-) 13https://github.com/crawl/crawl/commit/16c7f17c50ea 10:40:59 03DracoOmega02 07* 0.32-a0-1774-gdfbe08ab73: Place the Orb of Zot on an untransformable dais 10(7 minutes ago, 10 files, 25+ 16-) 13https://github.com/crawl/crawl/commit/dfbe08ab7398 10:44:18 <04d​racoomega> I am mildly amused that basically the first vault I've touched (aside from monster set changes or item renames, etc.) is the orb vault itself (and also mildly terrified that I have somehow messed up the syntax in some non-obvious way with Index not awake to double-check it for me :P) 10:45:19 <03i​mplojin> surely we'll find out if it's been broken when players complain in a few hours 10:45:44 <04d​racoomega> Hopefully not! 😛 10:46:20 03DracoOmega02 07* 0.32-a0-1775-gf20c09c4ad: Remove a now-unused include 10(11 seconds ago, 1 file, 0+ 1-) 13https://github.com/crawl/crawl/commit/f20c09c4ada7 10:49:02 <03i​mplojin> @asciiphilia that's been a longstanding issue with webtiles and i'm not sure if it was ever fully fixed (a quick look at CUE's lobby shows several players with thousand hour idle timers), advil or gammafunk might know more (see commits 2a3aa404fa, 33df2375a9) 11:23:59 <08n​icolae> taking a quick eyeball at the commit, looks fine to me 11:24:31 <04d​racoomega> Yeah, it's simple enough that surely it's hard to screw up (but somehow it may also be literally the first KFEAT line I've ever written :P) 11:24:53 <08n​icolae> you're in luck, as i have written at least two or three KFEAT lines 11:24:58 <04d​racoomega> At least 😛 11:25:11 <08n​icolae> well, i was being modest. maybe closer to ten 11:25:44 <04d​racoomega> (I am totally planning to great Index when she wakes up by telling her, without context, that I just pushed a change to the orb vault) 11:25:55 <08n​icolae> hell yes 11:26:24 <08n​icolae> tell her that it came to mind after a discussion with nicolae 11:27:30 <08n​icolae> she will spring to look at the recent commits with an alacrity usually only seen in and around burning buildings 11:28:48 <04d​racoomega> Ahahaha 12:07:36 <09g​ammafunk> hrm, yes, I have this same issue 12:07:46 <09g​ammafunk> let me go actually check cdi's configuration 12:07:50 <09g​ammafunk> I had assume I have something out of date 12:16:26 <12a​sciiphilia> If we know the specific conditions under which it occurs, it would be easier to debug. It seems to happen intermittently. 12:18:11 <12a​sciiphilia> If the timer wasn’t working at all, I think there would be even more super idle users than this. Maybe it doesn't work in certain situations. Of course, this is just a guess. 12:18:19 <09g​ammafunk> sorry, I've now gotten distracted by the fact that my config.py is so old and out of sync with the repository version, due to the fact that it came from what's in the dgamelaunch-config repo 12:18:25 <09g​ammafunk> so I'm resyncing them which will take a bit 12:29:00 03Implojin02 07* 0.32-a0-1776-g44e0ba907c: fix: don't hide the orb in console 10(43 seconds ago, 1 file, 1+ 0-) 13https://github.com/crawl/crawl/commit/44e0ba907cfe 12:42:58 03DracoOmega02 07* 0.32-a0-1777-g83d8b1f7ab: Mark gems as critical items (ge0FF) 10(4 minutes ago, 1 file, 2+ 2-) 13https://github.com/crawl/crawl/commit/83d8b1f7ab4e 12:42:58 03DracoOmega02 07* 0.32-a0-1778-g5f6dd93176: Remove an unnecessary use of item_def::is_critical() 10(2 minutes ago, 1 file, 1+ 2-) 13https://github.com/crawl/crawl/commit/5f6dd931766b 12:42:58 03DracoOmega02 07* 0.32-a0-1779-ga363cd56c6: Don't ever claim the orb dais is '(summoned)' (ge0FF) 10(24 seconds ago, 1 file, 6+ 0-) 13https://github.com/crawl/crawl/commit/a363cd56c652 12:46:44 04Build failed for 08master @ a363cd56 06https://github.com/crawl/crawl/actions/runs/10065345041 12:49:24 03dolorous02 07* 0.32-a0-1780-gb95f67303f: Unbrace 10(2 minutes ago, 1 file, 0+ 2-) 13https://github.com/crawl/crawl/commit/b95f67303f06 12:51:34 03DracoOmega02 07* 0.32-a0-1781-g1cace68bb9: Give the orb dais a proper feature type name (ge0FF) 10(25 seconds ago, 1 file, 2+ 0-) 13https://github.com/crawl/crawl/commit/1cace68bb91e 12:52:39 03dolorous02 07* 0.32-a0-1782-gb48ef885c9: Fix typo. 10(32 seconds ago, 1 file, 1+ 1-) 13https://github.com/crawl/crawl/commit/b48ef885c943 12:54:04 <04d​racoomega> Damnit T.T 12:54:19 <04d​racoomega> That was literally the entire commit 😛 12:54:41 <04d​racoomega> (I misspelled the feature itself in the same way at first, though at least I double-checked that before pushing it >.>) 12:54:53 <06d​olorous_84348> Happens 🙂 12:54:55 <12a​sciiphilia> I suggest some minor updates. - Add CNC to FooTV (Wouldn't it be better to add CDI as well?): https://github.com/crawl/dcss_tv/pull/1 - Update Wiki - DCSS Servers overview (CDI is also missing): https://github.com/crawl/crawl/wiki/DCSS-Servers-overview | CNC | [ubuntu 24.04 noble LTS](https://hub.docker.com/_/ubuntu) | gcc 13.2.0 | ✔ YES | No chroot, Containerized | - The CNC server link is missing from the 4. Playing Online 12:54:56 section on the right side of the dev blog. - Tournament page before 0.22 cannot be accessed. URL: http://dobrazupa.org/tournament/0.22/overview.html 12:57:18 <09g​ammafunk> Regarding 4., those files were lost when the host had a disk failure so we only have internet way back machine, sadly 12:58:01 <12a​sciiphilia> I'm glad it was backed up on the Wayback Machine. 12:58:31 <09g​ammafunk> Perhaps they could be downloaded from wayback machine to make a better archive 12:58:44 <09g​ammafunk> Not sure how easy that is to do 12:59:20 <12a​sciiphilia> https://web.archive.org/web/20190103154224/http://dobrazupa.org/tournament/0.22/overview.html 12:59:56 <09g​ammafunk> I can certainly put a copy on CDO but I need the files in a hostable form 13:02:24 <12a​sciiphilia> If you download it as MHTML and then convert it back to HTML, you should be able to do something similar. 13:02:27 <12a​sciiphilia> https://private-raw.abstr.net/crawl-0.22.html 13:06:06 <12a​sciiphilia> If you're planning to preserve all the possible hyperlink pages, it could be quite a tedious task. It would have been nice if Wayback Machine provided an API. 13:11:37 <12a​sciiphilia> Upon checking, it seems that we can access up to the 0.13 tournament on the Wayback Machine. 14:01:38 at some point it becomes easier to just point to the wayback machine… 14:10:44 <09g​ammafunk> Yeah it feels only worth archiving if we can preserve all hyperlinked pages generated and have all links work. Presumably this can be done with some kind of spider but it's non-trivial in my limited experience. Just a main link point to the way back machine seems best if there's no reasonable comprehensive solution 14:13:07 <09g​ammafunk> I need to ask greensnark a out this after I look into some things for CDI. CNC is probably ready to add now but I'll wait till this weekend when I can ask him about both servers for footv 14:13:46 <09g​ammafunk> For cdi I want to get the S3 setup working along with the S3 listing script 14:20:26 03mgdelmonte02 07https://github.com/crawl/crawl/pull/2677 * 0.32-a0-1783-g334656ff87: Add `crawl.stack` function to return the current Lua stack. 10(1 year, 11 months ago, 1 file, 26+ 0-) 13https://github.com/crawl/crawl/commit/334656ff8732 14:39:48 <09g​ammafunk> Nice 14:40:10 <09g​ammafunk> Going to merge this soon after I check it out 14:48:29 <12e​bering> that seems like one that we might not want to have available on public servers? 14:48:55 <12e​bering> (not actually a security expert, just afeared for memory-introspection-facing-the-public is bad cargo culting) 14:49:39 <03i​mplojin> vaguely skeptical that this is a good idea or needed in a clua context but what do i know 14:58:53 <12a​sciiphilia> Is there a specific reason for storing it in S3? Is it because it's cheaper for storing ttyrec files? 15:01:11 <12a​sciiphilia> One concern is that my server periodically compresses ttyrec files using dgl compress-ttyrecs. I'm not sure if compressed ttyrec files are supported. 15:09:14 <09g​ammafunk> Yeah, S3 is cheaper storage by far. My ttyrecs are compressed on cdi (via a batch job that runs daily), but I'm just past 50% full partition now 15:09:45 <09g​ammafunk> and you can get several hundred gigabytes of storage from S3 providers for very little 15:11:21 <09g​ammafunk> CDI is just a "droplet" VM from digital ocean that comes with 4 vcpus and 160GB of disk 15:11:39 re the stack function: I can see some possible uses in debugging recursion-loops in clua, but I'm still a bit dubious 15:11:51 (and the `sprintf` worries me. I commented) 15:11:59 <03i​mplojin> i mean there are what, like two downstream users of this? qw and whatever mgdelmonte is doing 15:12:03 <09g​ammafunk> thanks for looking at it 15:12:14 <03i​mplojin> surely it would help with clua bot things but 15:12:43 <09g​ammafunk> well, there's no problem at all with adding clua specifically to help with e.g. qw 15:13:04 <09g​ammafunk> we don't really care about minimizing complexity in that regard, since clua stuff is reasonably straightforward 15:13:08 maybe we need blua that's not enabled on servers 🙂 15:13:15 <09g​ammafunk> but of course if the implementation has security issues, then it's a problem yeah 15:13:37 <09g​ammafunk> yeah, that would be regrettable, to have to have a compilation flag to enable 15:13:53 <09g​ammafunk> hrm, it could work actually, if it was only allowed in clua if crawl were launched with a specific arg 15:14:12 <09g​ammafunk> because I have qw flagged as a bot user and it already launches with special arguments via modifications I've made to the dgl launcher script 15:14:21 <09g​ammafunk> currently to disable lua throttle and give clua more memory 15:14:41 <09g​ammafunk> so conditionally enabling something could be fine so long as there's good control over it 15:15:05 <09g​ammafunk> and clo is a fine way to control that for servers 15:15:29 <09g​ammafunk> but specifically what I need is a way to get a backtrace when there's a lua runtime error in qw 15:15:49 <09g​ammafunk> and with a possible additional complication that I'm running the main qw code in a coroutine (for a couple reasons) 15:16:14 <09g​ammafunk> not sure if that PR actually gives me what I need for such errors, would need to try it 15:16:54 <09g​ammafunk> however for people making elaborate clua things in general, could be extremely useful (and people do develop those things locally where it's ok to launch crawl with a special argument) 15:17:22 <09g​ammafunk> can even do what we currently do with the lua throttle and allow e.g. insecure stacktrace by default for non-dgl builds 15:17:40 <09g​ammafunk> and require a specific argument to enable for dgl builds (this is the current setup for the lua throttle) 15:18:21 <09g​ammafunk> as an aside, I don't actually know what lua things mgdelmonte is working on since I'm not sure if it's ever been posted 15:18:28 <09g​ammafunk> presumably not qw things 15:27:30 <09g​ammafunk> I think footv is just fine with gzipped ttyrecs, you should be ok there 15:39:06 Unstable branch on underhound.eu updated to: 0.32-a0-1782-gb48ef885c9 (34) 15:56:47 03regret-index02 07* 0.32-a0-1783-g7af4bb0ebb: New Xom action: friendly + hostile clone drop 10(2 days ago, 5 files, 99+ 6-) 13https://github.com/crawl/crawl/commit/7af4bb0ebbbe 15:56:47 03regret-index02 07* 0.32-a0-1784-g6311a4899b: New Xom action: wave of despair 10(2 days ago, 7 files, 92+ 3-) 13https://github.com/crawl/crawl/commit/6311a4899ba8 15:56:47 03regret-index02 07* 0.32-a0-1785-gd06570b3e2: New Alchemist starting-screen icon (Darby) 10(30 seconds ago, 3 files, 2+ 2-) 13https://github.com/crawl/crawl/commit/d06570b3e261 16:07:31 03mgdelmonte02 {GitHub} 07https://github.com/crawl/crawl/pull/2677 * 0.32-a0-1784-g1e909ee25c: snprintf 10(36 seconds ago, 1 file, 1+ 1-) 13https://github.com/crawl/crawl/commit/1e909ee25cda 16:09:40 03dolorous02 07* 0.32-a0-1786-g0c0b0355b8: Fix spelling. 10(59 seconds ago, 1 file, 1+ 1-) 13https://github.com/crawl/crawl/commit/0c0b0355b8e4 16:22:54 <12a​sciiphilia> I once used S3 for around 50GB of personal music/video storage, and I was surprised by the high costs, including transfer fees, which amounted to $20-30 (monthly fee). Because of this, I couldn't determine if it was cheaper than increasing disk storage in the cloud. However, with compressed ttyrecs and S3's dynamic pricing, it should definitely be more affordable. Perhaps, even without using dynamic pricing, the cost might still be 16:22:55 manageable. 16:49:51 <09g​ammafunk> Yeah, the S3 on digital ocean is $5/mo for 250GB, for example 16:50:09 <09g​ammafunk> and 1tb/mo outbound transfer 16:51:29 <09g​ammafunk> for block storage on DO you only get 100GB for $10/mo 22:35:31 Unstable branch on crawl.develz.org updated to: 0.32-a0-1786-g0c0b0355b8 (34) 22:58:40 Windows builds of master branch on crawl.develz.org updated to: 0.32-a0-1786-g0c0b0355b8 23:34:23 Unstable branch on cbro.berotato.org updated to: 0.32-a0-1786-g0c0b0355b8 (34) 23:55:21 Monster database of master branch on crawl.develz.org updated to: 0.32-a0-1786-g0c0b0355b8