00:00:12 Unstable branch on CRAWL.XTAHUA.COM updated to: 0.24-a0-115-gd155597 (34) 00:41:42 -!- TAS-2012v is now known as TAS_2012v 01:27:43 Unstable branch on crawl.develz.org updated to: 0.24-a0-115-gd155597 (34) 01:59:39 Windows builds of master branch on crawl.develz.org updated to: 0.24-a0-115-gd155597 02:59:06 shouldn't the first target selected when casting a spell be the highest HD monster ? like tab does ? 02:59:32 Monster database of master branch on crawl.develz.org updated to: 0.24-a0-115-gd155597 03:03:06 <|amethyst> would be nice if they shared logic, especially since autofight monster prioritization is customizable with lua 03:03:42 <|amethyst> though maybe it would make sense for them to prioritize different things 03:04:29 <|amethyst> (by default) 03:04:44 =nemelex SpSk 03:04:46 Defined keyword: nchoice => GhVM|TeEn|VpCj|VpAE|CeAE|DDIE|DEMo|FeWr|HaEE|GnHu|KoAK|FoSk|DDFE|OpAK|SpSk 03:05:30 <|amethyst> though I guess threat level (it doesn't use HD directly) is one of the *last* tiebreakers for tab with the default priority 03:05:51 <|amethyst> !source compare_monster_info 03:05:52 1/2. https://github.com/crawl/crawl/blob/master/crawl-ref/source/dat/clua/autofight.lua#L226 03:18:01 I have some macro like "za." and i think i'm going to remove the "." because it's too dangerous 03:31:58 Fork (bcrawl) on crawl.kelbi.org updated to: 0.23-a0-1729-gdd55c3d1e3 06:33:06 How long does it take for changes to rc to show up? Yesterday I changed my clan name an hour before the freeze date, and it's still hasn't come through. 06:34:06 s/it's/it/ 06:46:02 Shafts collapse without shafting monster 13https://crawl.develz.org/mantis/view.php?id=11882 by Le_Nerd 13:43:58 Yermak: it probably should have applied, what was the change you made? 13:48:56 floraline: can you fix messaging on CKO? https://cdn.discordapp.com/attachments/205316046230388737/546402529479491594/unknown.png 13:49:42 elliptic: my undelivered console message because CKO console messaging is broken: "that is not a lot of int you have" 13:50:52 why does console even have a completel separate messaging system 13:51:41 because console's came first and console doesn't speak compressed json 13:52:19 well, it does, there's two-way message passing 13:52:25 but "came first" seems likely 13:57:43 yes, iirc it's two-way now because the webtiles daemon was modified to know how to "speak" console messaging 13:57:50 which came late; they used to be disconnected 13:58:06 we also have DO 13:58:09 er CDO 13:58:14 which I'm not sure is built with webtiles 13:58:33 it's not running the webtiles daemon, certainly 13:58:42 right 13:58:56 gammafunk: who needs to be intelligent 13:59:07 yeah I guess you had exactly enough int 13:59:24 elliptic: any general thoughts on 0.23 now that you've had a chance to play it? 13:59:32 was a little scared of oof mutating -int, but 13:59:51 I was sort of waiting for the kown pie effect too, yeah 13:59:59 *klown pie 14:00:19 klown pies weren't too bad for me, but it also did mean that klowns were a real enemy that I tried to fight one on one etc 14:00:43 one comment is that making runed doors glass all over the place is annoying for stealthy characters 14:01:23 since it makes it very likely that the stuff behind the glass runed door will be awake by the time you actually consider opening it 14:01:41 yeah, that's come up 14:02:10 I wonder if it would make sense to have stealth checks blocked by transparent features? 14:02:10 I was just thinking that, yeah 14:02:10 maybe a bit abusable 14:02:11 well, or not really abusable 14:02:20 it's manipulable, but not necessarily in a bad way 14:02:22 I'm unsure 14:03:15 about klowns, I haven't bothered looking up what they actually do now, but I think the most annoying thing they've done so far is polymorph another monster into a doom hound 14:03:36 and then I didn't know what doom hounds did, so... 14:03:42 heh 14:03:49 nice 14:03:50 that's a good poly 14:03:58 I just got some boring -potion stuff 14:04:07 oh and one canceled a tele one time 14:04:17 hrm 14:04:23 I don't think that's actually a klown effect 14:04:29 actually no, that was a pan lord with chaos brand 14:04:29 they don't have dispel nor tele other 14:04:47 oh, yeah that's a pretty rare one 14:04:58 although maybe the disto tele effect also cancels? 14:05:08 not to say that -potion is boring, it was just boring for this char 14:05:13 yeah, tele should always cancel tele 14:05:59 mainly 0.23 just feels like any other version 14:06:16 new trap system seems to work well (though possibly the alarm trap frequency is a bit high) 14:07:08 !lm elliptic br.enter=gauntlet 14:07:29 3. [2019-02-15 20:13:22] elliptic the Grappler (L12 TrAK of Lugonu) entered a Gauntlet on turn 13461. (Lair:2) 14:08:14 I'm afraid I was lame and left my gauntlet right away 14:08:19 gauntlets also a place where glass plays weirdly with stealth 14:08:34 yeah, those a glass to the max 14:08:48 part of the continuing tradition of stealth nerfs 14:09:22 I sort of dislike how you can look several steps ahead in gauntlet at the start 14:09:42 would prefer only being able to see the very next options 14:10:03 yeah, that's certainly a possible route, but if that was done, I'd want exits at each step 14:10:30 seems reasonable 14:11:07 it is sort of just a hassle to go peering at each vault even if it's fun (to me) to plan a route through 14:24:28 @??redback 14:24:28 redback (04s) | Spd: 15 | HD: 9 | HP: 22-32 | AC/EV: 2/12 | Dam: 1804(strong poison:33-58) | cling, web sense | Res: 06magic(20) | Vul: 09poison | Chunks: 14noxious | XP: 220 | Sz: tiny | Int: animal. 14:41:41 gammafunk: for some reason dgl is failing to create player spool files. if they already exist it will write into them just fine, but it won't create them and i can't figure out why 14:42:16 hrm, not sure why that would be 14:42:31 directory permissions? 14:42:37 that would be the most likely culprit 14:42:58 i can change to crawl user and create them myself so i don't think that's it 14:43:47 perhaps it has a security thing where it refuses to create them for some reason; not sure if that had been handled by a setup script in the past 14:49:03 looking at what dgl does it's fopen with "a" which should create the file if it doesn't exist 14:51:55 chmod +t on the spool dir? 14:52:08 although if it's all the crawl uid then no 14:52:17 ...hm. selinux enabled? 14:54:06 (or systemd using cgroups, which can cause all sorts of hard-to-diagnose "fun") 14:54:31 yeah cdo has a different setup 14:54:36 with a dgl user and a crawl user 14:55:19 so dgl owns the spool dir with crawl as the group and with group being read-only 14:55:55 <|amethyst> gammafunk: dgl on CDO is more or less like crawl on most servers, and crawl on CDO is more like crawl-dev elsewhere 14:56:06 ah, good to know, thanks 14:56:14 <|amethyst> (CAO uses 'games' rather than crawl) 15:01:16 i fixed it but i don't understand why it wasn't working before 15:01:28 i guess i had the permissions wrong 15:02:05 thank goodness elliptic can finally read my life-saving console messages again 15:02:08 thanks! 15:02:15 my spool dir is /var/mail in chroot owned by root:mail 15:02:28 crawl is in mail group so it should have been able to write there, and indeed i could create files when i change to crawl user 15:02:44 but dgl running as crawl could not create files there 15:03:29 floraline: is crawl in the mail group according to the chroot? 15:03:37 or just according to your system 15:04:29 I'm not sure how user mappings precisely work going from system to chroot, but maybe there's a mismatch there 15:04:30 yes, it should be 15:09:52 now i just have to figure out why console games sometimes don't get found by webtiles 15:13:41 for posterity: i solved it by giving /var/mail in chroot to crawl. this will surely not cause any side effects 15:16:54 floraline: well that's essentialy how it's set up on CDO, so you should be fine 15:17:28 floraline: when you said you became user crawl, did you do so inside the chroot? 15:17:46 yes 15:17:47 by inside I don't mean cwd but rather under chroot 15:18:00 weird 15:18:04 sudo chroot xx && su crawl 15:18:13 huh, welp 15:18:36 but yeah, as |amethyst said, for cdo the dgl user is equivalent to your crawl user 15:18:56 and on cdo dlg is the owner of that spool dir inside the chroot 15:19:01 s/dlg/dgl 15:20:57 floraline: Oh, thanks, I'd noticed that was broken and failed to fix it. (I added "crawl" to group "mail", which should work but didn't and I dunno why) 15:21:13 that's exactly what i tried to do 15:22:36 there must be something weird about the way dgl runs as crawl that doesn't grant it exactly the same permissions 15:33:39 Except such a thing isn't possible (I would have said) 15:36:10 Oh, no, hang on; it setuid-s to "crawl", but doesn't setgid (or doesn't setgid to "mail", anyway). 15:40:10 i see. so dgl does not get write permission to the spool dir through a group. it must be owned by your dgl user, which is crawl if you are following the setup guide 15:40:56 that would be worth adding to the guide 15:44:33 Vexingly I fixed this once already and forgot; I've a feeling a Debian package changed it back (maybe next time I'll remember) 16:03:40 Pinkbeast, there's a reason I mentioned selinux and systemd/cgroups 16:03:43 both can do that 16:10:28 True, but I've got neither and I think "it doesn't setgid" explains it 16:18:15 yeh 16:34:51 advil, "Mercenary" -> "Mercenaries", also added a couple of teammates (they also have the required line in rc). 16:35:33 Yermak: did you add that before the clan cutoff? 16:35:41 an hour before that 16:38:07 Yermak: looks like you didn't update CKO 16:38:11 which has precedence 16:38:26 that is described on the rules page 16:38:46 I don't think anyone would strongly object to allowing the membership change 16:38:54 but that's something to be aware of when changing your team 16:39:04 ah, yes, it's still reading from https://crawl.kelbi.org/crawl/rcfiles/crawl-0.23/Yermak.rc 16:39:32 so unless advil objects, we can just modify your CKO rc (or delete it) from the clan stuff...although 16:39:35 yeah, seems reasonable to let it through 16:39:43 I'm not sure if the script will actually process it 16:39:47 as it does internally know about the cutoff 16:39:59 oh damn, right, I copied my rc to CKo in the middle of the tournament 16:41:04 advil: I guess we could temporarily change the deadline 16:41:10 restart the script and have it run a one-off 16:41:40 since the rcfile fetching was properly turned off (admitedly I did it a bit late) 16:41:45 if Yermak's fetched CKO rc is deleted 16:41:49 it should work 16:41:56 then we can change it back 16:42:09 I don't think that would lead to any spurious clan changes since the fetching was turned off 16:42:28 but it might let some people sneak changes through if they made them after the deadline but before I turned off the fetching 16:42:37 which was like 4h after deadline 16:42:42 that sounds fine to me, but yeah, I wonder if this affects anyone else 16:42:43 it's sort of easy to forget to do that 16:42:49 I've modified CKO rc. 16:42:55 I didn't even know it had to be turned off manually :) 16:42:59 Yermak: doesn't matter, we're not fetching them any more 16:43:05 ah 16:43:21 I don't have a great way of knowing if any clan defs would change 16:43:25 aside from yermaks 16:43:39 I suspect not since it'd have to be what I described in terms of the timing 16:44:41 if advil wants to brush off his SQL-foo and check that somehow, or we can just do it and see if anyone complains 16:45:27 I think that query might be a little difficult to make though 16:45:39 looks like the script uses its run time to compare against the deadline 16:45:50 right 16:46:14 so my plan was to change deadling to like "now + 30 minutes", stop it, run a one-off 16:46:21 then change it back and restart the daemon 16:46:30 well we'd stop the daeamon before doing this 16:46:59 ah I see 16:49:29 oh, what you turned off is whatever is fetching the rc files? 16:50:42 my knowledge of these scripts is still extremely weak 16:52:18 advil: yes, it's just some crontab entries 16:52:25 it seems like that happens with a cronjob, if that cronjob is no longer running, then manually updating that one rcfile (deleting or fixing rcfiles-cko/Yermak.rc), and then rerunning the team update listener thing, seems like it should work 16:52:29 those are what fetch RCs (or in the case of CAO, a text file) 16:52:32 maybe that's what you already said :) 16:52:37 it is, yes 16:52:50 well you also have to modify the deadline 16:52:52 in the config 16:52:55 but yeah basically that 16:52:58 yeah 16:53:10 ok, and you were worried about any rc changes that happened in those 4 hrs 16:53:25 yes, since I turned off fetching 4 hours after 16:53:26 I'm ok with letting them have it 16:53:30 ok 16:53:33 I'll do that now then 16:53:37 I don't think anyone would get upset about this? 16:53:54 maybe I'm underestimating what people could get upset about 16:54:10 yeah, I suppose theoretically if someone was lucky to sneak in a clan change and someone else saw that 16:54:17 said observer might raise an objection 16:54:27 the thing is we wouldn't really know about it unless someone brought it to our attention 16:54:32 so I don't think it's a big concern 16:54:33 yeah 16:54:57 we could also effectively revert that change but doing something similar to what we're doing with Yermak 16:55:43 I think also if we're going to grant yermak an exception for this kind of case (where there was a misunderstanding about which rc to update, since it was basically a technical/interface issue) if someone else asks we would need to do it again 16:56:35 sure, but I also think we might need to have a "discretion is left to tournament organizers" clause 16:56:52 maybe that section of the tournament rules could have a note 16:56:52 that's a good feature to add in the future, an explicit indicator of which rc file it is reading, since that info looks to be easy enough to get at 16:57:04 yeah 16:57:06 oh yeah, you mean on the team page? 16:57:08 that would be cool 16:57:12 or the player page, or both 17:07:23 !time 17:07:24 Time: Feb 16, 2019, 10:07:24 PM, UTC. The 2019 0.23 tournament ends in 7 days, 21 hours, 52 minutes and 35 seconds. 17:10:45 Yermak: ok, should be updated 17:13:34 thank you 17:13:56 also is this some kind of crazy new alt-based tournament strat? 17:19:06 I'm just wondering if yermak's mugl will make it 17:19:25 only 22 hp next to a blind orb of fire and a very angry ancient lich 17:19:35 but it's parked for now 17:19:53 and fortunately not part of the main streak 17:28:14 heh, did you watch it? 17:28:57 Plan is to kill oof with /scattershot, survive a turn against alich, then read silence. 17:31:53 Yermak: yeah I did, since I eventually got so curious as to why it was parked so long 17:32:26 right, I was expecting silence maybe earlier; I probably didn't pay quite careful enough attention, but I think the orb was blinded 17:32:38 however if you kill the lich first the orb is still more likely to fireball you 17:34:53 I know that for sure 17:36:10 If I palyed more carefully I could've used cheesy strategy based on "Everyone hates you": spawn hostile water elemental behind Me, so that oof wouldn't cast even bolts of fire. 17:37:16 we've talked about nerfing those sac loves strats by making the evokers either just never make summons or disallowing them 17:37:48 I think it would require a ruling from the Ru overlord Lasty to decide that, though 20:00:02 fake abjuration vs normal abjuration differences are very weird 20:13:01 geeze this is just a byzantine nightmare; the notiion of what is a summon and what isn't just isn't very coherent 22:31:18 FR: issue a warning upon casting Summon Forest with OTR active.