01:08:32 Stable branch on crawl.develz.org updated to: 0.25.0-37-gb73750e0fb (34) 01:33:02 ebering: would you accept a pr to remove food shops? 01:34:07 now that food is less overloaded, the food clock should be able to be tighter. Food shops add a lot of potential variation to the clock length; and when food isn't required (majority of the time) they are very boring 01:38:08 Unstable branch on crawl.develz.org updated to: 0.26-a0-249-g7253cd193c (34) 01:53:01 Forgeling (L27 BaAM) Crash caused by signal #6: Aborted (Zot:3) 02:27:42 -!- amalloy is now known as amalloy_ 03:19:42 Stable (0.25) branch on crawl.beRotato.org updated to: 0.25.0-37-gb73750e 03:32:39 Unstable branch on crawl.beRotato.org updated to: 0.26-a0-249-g7253cd1 (34) 03:46:18 Stable (0.25) branch on crawl.kelbi.org updated to: 0.25.0-37-gb73750e0fb 03:47:45 Experimental (delpal) branch on crawl.kelbi.org updated to: 0.26-a0-241-g4e6e3e941d 03:52:37 Fork (bcadrencrawl) on crawl.kelbi.org updated to: 0.22.1-2838-g272d1e3184 09:00:42 alexjurkiewicz: see the earlier discussion in -dev with elliptic 09:02:08 probably going to remove food outright. I'll look into that today and maybe tweak the clock to make trunk more playable if it will take a while 09:02:33 so removing food shops in the interim is a piece of the puzzle and appreciated to weave into the outright food removal 09:14:42 ebering I have an in-progress branch fyi 09:14:58 it's more involved than I expected (and I expected it to be more involved than I expected) 09:15:58 advil: if you want help you can stick it in the repo 09:22:33 03ebering02 07* 0.26-a0-250-g75e6479: Double food generation 10(82 seconds ago, 2 files, 5+ 5-) 13https://github.com/crawl/crawl/commit/75e647984665 09:23:09 ^ says in the commit message that this is a stopgap 09:23:11 worth noting that if food shops are replaced by other shop types then that is effectively a straight player buff 09:24:37 I don't know offhand what proportion of shops food shops are, but might be reasonable to decrease shop generation by the same amount 09:26:17 of course, removing hunger is also a player buff (because it sometimes caused players spamming high level spells or ely invocations or the like to need to take a turn off in combat to eat) but that one is harder to quantify or compensate for 09:28:28 ebering a variant of that I was considering was doubling nutrition per food 09:28:35 which affects a bunch of other cases 09:28:44 though as a stopgap probably doesn't make much of a difference 09:29:01 (maybe worth mentioning that my impression is that the addition of shop vaults containing 3 or 4 shops also was a straight player buff? but I haven't actually looked at shop placement code in years) 09:32:14 Unstable branch on crawl.kelbi.org updated to: 0.26-a0-250-g75e6479846 (34) 09:32:21 I'm not sure there's very many of those or that they're very recent, but they can lead to extreme shop spikes if they place together with serial_shops 09:33:59 they aren't recent yeah and I don't know how significant they are, but my impression is that they count as just 1 shop for generation purposes and thus are some sort of increase 09:33:59 yeah, that's correct 09:34:01 probably food shops getting replaced by other shop types is more significant 09:34:02 there was a case someone posted on reddit where I looked closely into the details, I can find the seed 09:35:48 https://www.reddit.com/r/dcss/comments/gtaggj/12_shops_on_d4/ they say 12 but it was only 9 -- basically shope_serial rolled its max (5) and combined together with nicolae_shop_statue_plus which places 4 09:36:35 *shop_serial 09:36:46 I had 10 on a level recently iirc 09:36:48 I would guess that 9-10 is about the absolute max 09:36:50 heh 09:37:05 possibly similar 09:37:16 probably something like nicolae_shop_statue_plus shouldn't actually place if serial_shops is there 09:39:30 set up luniq_shop or something 10:25:54 saddest is to remove honeycomb flavor :-( 10:25:58 also pan pizza 10:33:15 F 10:46:31 actually honeycombs -> ambrosia is ok in many places 11:13:29 New branch created: deflavour (2 commits) 13https://github.com/crawl/crawl/tree/deflavour 11:13:29 03advil02 07[deflavour] * 0.26-a0-242-gb403669: Fix a couple of minor formatting issues in INSTALL.md 10(14 hours ago, 1 file, 7+ 4-) 13https://github.com/crawl/crawl/commit/b40366909301 11:13:29 03advil02 07[deflavour] * 0.26-a0-243-g97f76ea: Remove rations 10(8 minutes ago, 57 files, 194+ 448-) 13https://github.com/crawl/crawl/commit/97f76ea0406c 11:13:59 oops no idea how b403669093013 got on that branch 11:15:03 03advil02 07* 0.26-a0-251-gb87c5ff: Fix a couple of minor formatting issues in INSTALL.md 10(14 hours ago, 1 file, 7+ 4-) 13https://github.com/crawl/crawl/commit/b87c5ff4f8df 11:17:06 Unstable branch on crawl.kelbi.org updated to: 0.26-a0-251-gb87c5ff4f8 (34) 11:20:11 03advil02 07[deflavour] * 0.26-a0-252-gabff4ab: Remove rations 10(14 minutes ago, 57 files, 196+ 449-) 13https://github.com/crawl/crawl/commit/abff4ab3234a 11:31:41 just to be clear I have no plans to unilaterally merge this or anything... 11:31:52 more of a contribution to the discussion 11:32:22 (and tbh sometimes I get curious how to do a thing) 11:32:24 04Build failed for 08deflavour @ 97f76ea0 06https://github.com/crawl/crawl/actions/runs/162113575 11:35:25 03advil02 07[deflavour] * 0.26-a0-253-g4eb5d95: Linting / test fixes 10(26 seconds ago, 2 files, 2+ 2-) 13https://github.com/crawl/crawl/commit/4eb5d95aeb17 11:46:03 04Build failed for 08deflavour @ abff4ab3 06https://github.com/crawl/crawl/actions/runs/162121774 11:53:43 04Build failed for 08deflavour @ 4eb5d95a 06https://github.com/crawl/crawl/actions/runs/162137647 11:55:32 03advil02 07[deflavour] * 0.26-a0-254-g950df13: Even more linting 10(18 seconds ago, 1 file, 0+ 2-) 13https://github.com/crawl/crawl/commit/950df1362a8f 12:36:35 perhaps its time for a new CI target: REMOVE_FOOD=YesPlease 13:10:56 Unstable branch on crawl.akrasiac.org updated to: 0.26-a0-251-gb87c5ff (34) 13:21:51 The build passed. (deflavour - 950df13 #13840 : advil): https://travis-ci.org/crawl/crawl/builds/706220797 13:27:47 -!- amalloy_ is now known as amalloy 14:28:42 advil: wow, the warnings when compiling dgl and virus are kind of scary in terms of how many they are 14:29:00 with a recent version of gcc, at least 14:29:35 haha yeah 14:30:34 also, I made a very dumb mistake of installing following package https://packages.ubuntu.com/bionic/docker, which is "System tray for KDE3/GNOME2 docklet applications" and I was rather confused for a bit when I didn't have a docker command 14:30:34 I think our warning ocd is a bit unusual, at least historically 14:31:14 I could see that 14:37:06 hrm, guess there's not a good way to pass the number of cpus to the make via -j 14:37:17 or maybe there is something I can edit in dgamelaunch 14:37:44 but I guess I probably don't want that for rebuilds since I only have two vcpus on this thing 14:52:50 boy, gonna be interesting to see if you can really run a server with any kind of usage with these resources 14:55:46 what are you running on? 14:58:58 just a lightsail 2vcpu, 4gb memory, 80gb disk instance 14:59:13 I believe that's roughly what cko is by default 14:59:14 I guess kind of analogous to the one floraline is using for cko 14:59:28 disk space will be a limiting issue there if this is to be a live server 14:59:34 right 15:00:11 I can certainly add a volume and/or delete ttyrecs according to the 'advil recommended ttyrec scheme' :) 15:00:38 but really this is a testing server for now, not official, so I just won't see too much usage I think 15:01:47 advil: do you have any quick recommendations for setting up a hostname? I saw you had made one under dcss.io 15:02:55 I do own dcss.io, could probably add a subdomain under there (though I've never done that outside my hosting company so I don't know what exactly is involved) 15:05:09 ok, I'll look into the IP reservation scheme lightsail/aws has a bit more and look into some available domain and dns options 15:05:52 gammafunk: the rebuild script is all screwy, i think it should work if you leave the page alone. some problem with nginx i can't figure out 15:06:18 floraline: well, what happened the first time actually is that the connection just got dropped early into the build 15:06:33 but as advil suggested, retrying did fix it thanks presumably to ccache 15:06:45 oh i see. i have no idea what that was about then lol 15:07:04 I was postulating that the cgi was writing to buffered output and nginx was timeint out the connection due to inactivity, but that probably doesn't make sense 15:07:23 you got the first part correct, it's buffering the cgi output which is what upgrading to buster will fix 15:07:45 idk what kind of timeout would come into play there, either way it's pretty dumb behavior 15:08:02 yeah, guess that explains why I got bursts of text; something was buffered 15:09:10 and yes that sounds exactly like what is cko is on, except i only have 60gb of storage 15:09:10 offloading morgues and ttyrecs is essential 15:10:27 how do you offload ttyrecs btw? 15:10:31 well, and morgues, I guess 15:10:36 in terms of where you store them 15:12:25 in dgamelaunch-config i modified compress-morgue.sh and compress-ttyrecs.sh to copy the archives to s3 storage (spaces on digitalocean) and delet from the local disk 15:12:49 using s3cmd. so whenever a file would be compressed, it is also moved to s3 15:13:02 those cgi scripts do stop of the user disconnects, at least on other servers 15:13:10 *if 15:13:54 i also updated retrieve-ttyrec.py in utils/ to show all files from s3 as if they were on the local machine 15:14:41 cool, sounds like what alexjurkiewicz uses 15:15:08 yeah, he copied me after i showed him what i was doing. his is probably implemented in a more maintainable way if you want to reference something 15:15:40 mine's as hacky as it gets so i don't recommend using it in production 15:16:21 yep, I will check in with him about that when I get to that point 15:16:49 but in general i think the idea is sound, at the very least storing these files on the server itself is kind of a non starter 15:17:08 there's also the looming issue of some of the log files growing forever 15:17:27 but that's a very slow pace iirc 15:18:24 I think advil wants to address that at some point as well; one of our many infrastructure wishlist items 15:19:15 if you mean the webtiles server logs, they are rotating using the python log library which should place a max on the size used 15:19:30 also now that we're talking about, i'm going to run out of room in my $5 s3 storage eventually and i'm not sure i can afford a third place to store old files. it would be good if we had some kind of very long term storage for very old files, if we want to keep them at all 15:20:12 yeah this is something I started working on and then got annoyed by and flipped back to "why don't we just delete" 15:20:12 oh, yeah, thought logs was referring to the logfile and milestones file 15:20:12 not the server log, i think it's the meta log files? 15:20:16 oh, those do grow indefinitely 15:20:20 the one that these bots like to read and that we're not allowed to touch 15:20:20 for webtiles logs those can rotate for sure 15:20:36 alexjurkiewicz has an issue with some longer-term ideas 15:21:00 it is possible in principle to cycle those and archive them somewhere else, but we don't have a systematic way of doing that 15:21:36 i.e. to take a specific logfile/milestone pair for version X on some server, put them somewhere else, and start anew 15:21:46 where the X that matters most is trunk 15:24:55 advil: a pretty common workflow that would have resulted in b403669 being on that branch is: implement&commit b403669; create branch; implement&commit 97f76ea; push the branch. as far as your local trunk is concerned, the branch differs by only 1. but to github, a new branch has arrived, with two commits not on its trunk 15:25:43 yeah I did do that doc commit while in the middle of working on deflavour, so that sounds right 15:26:07 basically where I'm at with long term archiving right now is that the metaformat involved in archiving ttyrecs is too annoying to want to deal with, but if it were less annoying I'd consider it 15:27:00 where an example of a "less annoying" metaformat would be if all .ttyrec + .ts files associated with a particular game were collected together in a tarbell with a reasonable game-based naming scheme for the game 15:27:30 "game-based naming scheme" could be the same one that morgues use 15:29:04 the headaches of getting ts files in the right place so that footv could actually work on some servers would make me happy to see such a scheme 15:29:22 I don't even understand how the .ts files are actually used though 15:29:39 me neither, but if you don't have them ttyrec seeking doesn't work well? 15:29:44 in footv 15:30:35 I think seeking by turn doesn't work, yeah; so -tv off a milestone, and I'm not sure sure what all will work for an lg query either 15:31:15 but I don't really know the details at all 16:05:03 -!- Adeon_ is now known as Adeon 16:22:08 ok team I've re-written my charms plans based on the previously received input and discussion 16:22:11 https://github.com/crawl/crawl/wiki/Charms-Overhaul 16:22:18 highlights: won't have two skills 16:26:03 03ebering02 07* 0.26-a0-252-gb32a1a6: Make lemures true demons 10(2 minutes ago, 1 file, 1+ 1-) 13https://github.com/crawl/crawl/commit/b32a1a619543 16:29:22 i have an ongoing trunk game taking up 1GB of ram, anyone know why that might happen? 16:29:33 hm, what game? 16:29:44 frimble 16:30:17 nothing immediately obvious, can you get a copy of a save? 16:30:35 they are pretty late game 16:31:58 i copied it 16:32:15 and send it to me somehow :-) 16:32:25 Unstable branch on crawl.kelbi.org updated to: 0.26-a0-252-gb32a1a6195 (34) 16:45:55 or you could try it yourself, the immediate question would be if memory usage replicates on load, the second question would be if there's anything very obviously distinctive in the game history that could be used to try to replicate it 16:49:50 sorry i'm dealing with work right now 16:49:56 i can send it over email 16:50:13 no problem 16:55:35 any issue with sending SIGHUP to this player? that's graceful iirc 17:11:12 they just won actually 17:11:27 so you should be ok now 17:11:38 but yes I think a sighup would be ok? 17:11:48 so far haven't replicated any unusual memory use 17:12:06 (which probably does meant that getting them to restart their game one way or another would work) 17:14:20 -!- amrqx is now known as yesno 17:14:23 !lg frimble -log 17:14:24 4873. frimble, XL27 FoFi, T:81355: https://crawl.kelbi.org/crawl/morgue/frimble/morgue-frimble-20200708-210951.txt 17:14:49 -!- yesno is now known as Guest71490 17:15:02 !lg frimble x=start,end 17:15:03 4873. [start=2020-07-08 04:46:44 [20200608044644S];end=2020-07-08 21:09:51 [20200608210951S]] frimble the Conqueror (L27 FoFi of The Shining One), escaped with the Orb and 15 runes on 2020-07-08 21:09:51, with 18950093 points after 81355 turns and 3:11:01. 17:15:21 they may have played through this in one long session, which wouldn't help 17:15:27 -!- Guest71490 is now known as equivoker 17:16:08 if there's a memory leak somewhere 17:16:09 or 1-2 sessions 18:25:23 Unstable branch on underhound.eu updated to: 0.26-a0-252-gb32a1a6195 (34) 18:48:16 !lg frimble -2 -ttyrec 18:48:17 4873/4874. frimble, XL27 FoFi, T:81355: https://crawl.kelbi.org/crawl/ttyrec/frimble/ 2020-07-08.04:46:35.ttyrec.bz2 2020-07-08.19:48:57.ttyrec.bz2 18:48:25 ^ 2 sessions 19:06:43 +gammafunk | hrm, guess there's not a good way to pass the number of cpus to the make via -j <-- make -j $(nproc) you mean? 19:09:39 advil: pikel's slaves in x-v mention they provide no nutrition when killed. Which is something for your deflavour branch to address if it hasn't. I noticed this yesterday in trunk 19:12:10 Sonf of Slaying -> Hexes/Necro(!): Dirge of Death 19:12:33 how about spectral weapon as an artefat property?! 20:26:45 btw, i forgot to mention this earlier, but the barding merger increases boot generation by about 17%, if i did the math right 20:27:20 mundane/ego floor boots; vaults are obvo unaffected and randarts work differently 20:28:08 i'm not sure if we care and, if we do, what we'd do to change it 20:40:33 -!- misha is now known as werekitten 20:53:30 03PleasingFungus02 07[delpal] * 0.26-a0-270-g366e7ef: Merge remote-tracking branch 'origin/master' into delpal 10(28 minutes ago, 0 files, 0+ 0-) 13https://github.com/crawl/crawl/commit/366e7ef64168 20:53:30 03PleasingFungus02 07[delpal] * 0.26-a0-271-g71b175a: Palentonga: fix build 10(24 minutes ago, 2 files, 2+ 2-) 13https://github.com/crawl/crawl/commit/71b175a40ee6 20:53:30 03PleasingFungus02 07[delpal] * 0.26-a0-272-g9f53b1b: Palentonga: bard up (ebering) 10(3 minutes ago, 9 files, 36+ 30-) 13https://github.com/crawl/crawl/commit/9f53b1ba16e5 21:05:11 03PleasingFungus02 07* 0.26-a0-253-gcd7d158: Give Hedge Wizard a starting dagger 10(4 minutes ago, 1 file, 1+ 1-) 13https://github.com/crawl/crawl/commit/cd7d1585338d 21:05:11 03PleasingFungus02 07* 0.26-a0-254-gdaab700: Make Storm card electric (WiseOldWolf) 10(2 minutes ago, 1 file, 1+ 0-) 13https://github.com/crawl/crawl/commit/daab700853c7 21:12:15 04Build failed for 08delpal @ 9f53b1ba 06https://github.com/crawl/crawl/actions/runs/162656177 21:15:08 03PleasingFungus02 07* 0.26-a0-255-gd5f3108: minislime_mu tweaks 10(4 minutes ago, 1 file, 1+ 2-) 13https://github.com/crawl/crawl/commit/d5f3108e62e6 21:17:09 Unstable branch on crawl.kelbi.org updated to: 0.26-a0-255-gd5f3108e62 (34) 21:43:18 03PleasingFungus02 07[delpal] * 0.26-a0-273-ge9c3666: Palentonga: shift polearms apt down (Nomi) 10(55 seconds ago, 1 file, 1+ 1-) 13https://github.com/crawl/crawl/commit/e9c366633bb6 21:45:34 03advil02 07[deflavour] * 0.26-a0-255-g14370c8: Update Jiyva text for no rations 10(19 minutes ago, 3 files, 13+ 7-) 13https://github.com/crawl/crawl/commit/14370c8cb21c 21:45:34 03advil02 07[deflavour] * 0.26-a0-256-g26885d3: Fix a bunch of hints mode stuff for food 10(3 minutes ago, 1 file, 36+ 35-) 13https://github.com/crawl/crawl/commit/26885d3459d3 21:45:34 03advil02 07[deflavour] * 0.26-a0-257-g750db15: Remove a stray reference to nutrition (gammafunk) 10(34 seconds ago, 1 file, 2+ 5-) 13https://github.com/crawl/crawl/commit/750db1584123 21:48:30 The build is still failing. (delpal - 9f53b1b #13842 : Nicholas Feinberg): https://travis-ci.org/crawl/crawl/builds/706366632 21:59:00 04Build failed for 08delpal @ e9c36663 06https://github.com/crawl/crawl/actions/runs/162692750 22:35:04 The build is still failing. (delpal - e9c3666 #13845 : Nicholas Feinberg): https://travis-ci.org/crawl/crawl/builds/706375292