00:08:28 apt-get worked, didn't have to do it in chroot though 00:10:04 not sure we have anything that cares yet. that might change with species etc. data in yaml 00:42:05 -!- amalloy is now known as amalloy_ 02:32:11 bearlyready (L10 MuCj) ASSERT(env.map_knowledge(pos).item()) in 'files.cc' at line 1383 failed. (D:9) 02:35:35 bearlyready (L10 MuCj) ASSERT(env.map_knowledge(pos).item()) in 'files.cc' at line 1383 failed. (D:9) 02:35:48 bearlyready (L9 MuCj) ASSERT(env.map_knowledge(pos).item()) in 'files.cc' at line 1383 failed. (D:9) 02:48:08 Monster database of master branch on crawl.develz.org updated to: 0.24-a0-284-g9fb9971 05:20:23 -!- TAS-2012v is now known as TAS_2012v 09:22:53 Stable (0.22) branch on underhound.eu updated to: 0.22.1-59-ge965b3aba3 10:21:02 !crashlog bearlyready 10:21:06 3. bearlyready, XL9 MuCj, T:8813 (milestone): http://crawl.akrasiac.org/rawdata/bearlyready/crash-bearlyready-20190505-063547.txt 10:50:00 something with gozag & gold detection on level change 10:50:21 I should just give in and put seeds in crashlogs 10:50:47 would probably make it a lot easier to replicate bugs like that 10:59:04 https://crawl.develz.org/mantis/view.php?id=11950 10:59:10 It could be relevenat. 10:59:18 relevant 11:23:38 Stable (0.23) branch on underhound.eu updated to: 0.23.1-50-ga50397a6bc 12:12:41 Yermak: thanks, I'd missed that bug 13:12:59 -!- TAS-2012v is now known as TAS_2012v 13:15:53 hello 13:24:22 bearlyready (L9 MuCj) ASSERT(env.map_knowledge(pos).item()) in 'files.cc' at line 1383 failed. (D:9) 13:43:11 Thanks for your help yesterday geekosaur https://imgur.com/a/iYplql2 :) 14:23:50 any word on beem? 14:37:50 Where do logs for dgamelaunch go? Mine is crashing upon actually entering the game. 14:38:24 pwnmonkey: Before I look, NB you can run dgamelaunch from the shell to see what it does, and that can help 14:38:36 Oh I am, no output. 14:38:44 Just returns to prompt 14:38:47 Anything handy under strace? 14:39:25 floraline: I've noticed a couple of times the kelbi webtiles lobby saying I'm idling even when I'm playing (on console) 14:39:55 obviously not a big deal or anything but I thought you might want to know 14:40:28 glaas: there's a known problem with console games sometimes not showing up correctly on webtiles 14:40:29 pwnmonkey: But 'dgldebug.log' in the chroot, by default. 14:40:43 i've spent a lot of time looking into it over the past few months but i haven't fixed it yet 14:40:59 you should be able to fix it by saving your game and reloadig it 14:41:45 if you notice that webtiles isn't showing your game correctly, try reloading your game a few times until it does 14:41:58 Would the strace upon launch be important? or at time of (p)lay 14:42:54 I also don't see that log file under chroot 14:43:12 floraline: ok, thanks 14:43:41 I'd strace the whole thing (use -f) and work backwards from the end of the strace output, on the basis it probably did something weird just before it crashed. 14:44:18 Gotcha, thanks. 14:48:27 There's stuff like this, but idk if its normal operation: stat("/root/.terminfo", 0x5579f3ba52c0) = -1 EACCES (Permission denied) 14:48:31 https://pastebin.com/6Wd5MFw2 14:49:31 It may be easier to run it with "-o foo" to capture the output to a file, FWIW 14:50:37 Ah TIL, much better. https://pastebin.com/1bAxd6gz 14:51:06 openat(AT_FDCWD, "/usr/lib/locale/en_US.UTF-8/LC_CTYPE", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory) looks curious 14:51:34 also looks like my permissions are wrong for that log file: 8062 openat(AT_FDCWD, "/dgldebug.log", O_WRONLY|O_CREAT|O_APPEND, 0666) = -1 EACCES (Permission denied) 14:51:36 let me fix that up. 14:52:14 that makes me think it's in the wrong place, doesn't seem like it should be going in / 14:52:23 It does go in / (in the chroot) 14:53:19 -!- amalloy_ is now known as amalloy 14:53:51 Does that mean the crawl users needs write access to it? currently the wiki says only dgldir and crawl-master need it. 14:54:10 And to explicitly not give it to / 14:54:35 -rw-r--r-- 1 crawl crawl 9708 May 5 17:14 dgldebug.log 14:54:52 pwnmonkey: Please clarify what you mean by " And to explicitly not give it to /" 14:55:59 From wiki before the chown -R crawl:crawl stuff " DO NOT chown the whole chroot: various programs and directories inside the chroot need to be owned by root or other users. " 14:56:04 Just wondering if that's why it can't write to / 14:56:18 pwnmonkey: Please don't do anything 14:56:42 That advice is correct. 14:57:14 But it doesn't need to be able to write to arbitary files in / , just to dgldebug.log, which is why it's owned and writeable by "crawl" 14:57:38 so i just 'touch' the dgldebug.log and give it ownership? 14:57:55 "If you want debug messages, also create the file dgldebug.log and make it writable for crawl. Warning, the file has a reputation of not being incredibly useful." 14:58:11 So yes. 14:58:12 Gotcha. 14:58:27 Okay, it output something to the logs. Just this one line: "purge_stale_locks dir open failed" 14:59:37 Did you chown the directories it _is_ suggested to chown? 14:59:50 (If so, strace again and pastebin) 15:00:12 I believe so: https://pastebin.com/mZ7UxsAx 15:04:07 Also, I'm not sure if it's important. I didn't do any of the webtiles stuff. 15:04:16 pwnmonkey: So please run it under strace again and pastebin the result. 15:04:34 You don't have to do the webtiles stuff to make it work. 15:04:37 That was done earlier, previous strace had those chown settings. 15:04:50 I can again though 15:04:59 pwnmonkey: I have seen no strace after you created dgldebug.log 15:05:11 Oh since that, kk. 15:05:52 https://pastebin.com/daXkUC5G 15:07:36 Is there a dgldir/inprogress ? 15:10:19 Yes but it didnt have crawl as it's owner unlike the other subdirs.. interesting. 15:10:42 Oh, and it's misspelled. Oops. 15:10:50 Then you didn't "su crawl" in "9. Create directories", or have mis-chowned after, so it's worth checking the other stuff created then 15:11:25 Fixed that: drwxr-xr-x 2 crawl crawl 4096 May 5 12:09 inprogress 15:14:06 OK, have you checked the other stuff created in "9. Create directories"? (Also in general it would be useful if you said explicitly "and it still doesn't work", and then produced another strace pastebin) 15:16:02 Yeah working on that now, only thing wrong from that step was that I had misspelled inprogress. (subdirs were there) New strace is much longer as it seems to be actually loading crawl now, but just kicks me back out to the dgamelaunch menu. Doesn't crash though. https://pastebin.com/qVmD1hSz 15:16:39 That's identical to https://pastebin.com/daXkUC5G 15:18:45 Weird. The strace was too long to copy from term so I scp'd it out and it was the old one. 15:20:05 Hit my limit on pastebin. https://0bin.net/paste/lAYtRmidFUKmZLk7#yzOaU2LaCWU6qgO4q39-oM4pZqVjEaZRR8Sn0FDb8SF 15:20:28 Sorry, Javascript. 15:20:40 Suggest you pastebin the last _n_ lines 15:21:34 Limit as in you can only upload 10 per day. 15:21:59 Might only check cookies though 15:25:57 I need to run to work though real quick. Once I'm back I'll find a suitable alternative to paste. Thanks for your help so far Pinkbeast. 15:26:30 It might be by then this firefox addon signing nonsense is sorted out and I can look at that 0bin paste 15:26:57 Oh are addons still down? 15:30:46 from what I hear, you enable Studies and there's one with a workaround/hotfix 15:31:26 floraline: I'm just about to break my promise not to rebase master, sorry, but I think you build at a radically different time of day so I hope it won't bite you 15:32:11 we'll see if the rebuild script handles it 15:33:30 I think you won't notice at all - if you don't pull in the ten minutes between commit A being pushed and it being replaced with A+B, you'll never know. It's only if you got version A that you have trouble. 15:35:39 is that how it works? 15:35:51 won't the repo history be permanently rewritten? 15:36:04 It will, but you'll never see the pre-rewrite version. 15:41:12 yeh, it only matters if you see the part of the history that gets rewritten 15:41:20 maybe force push is what i'm thinking of 15:41:29 I did a force push, yes 15:41:50 yes, again, only matters if you've seen that history 15:42:06 The history of the branch is (ideally) like pages in a book, one per commit. I pushed page 23, and it should never change. However, I rebased and force pushed, saying "actually here's a new page 23". 15:42:24 Now if you had pulled in that interval, you go "wait, I've got a _different_ page 23? WTF?" 15:42:36 But if you haven't you get the replacement page 23 and you're not the wiser. 15:42:54 oh so you're saying i haven't yet pulled the history which you've rewritten 15:43:08 so from my perspective nbd 15:43:25 that makes sense 15:43:33 I hope that's the case, yes, but I thought I'd mention it in case you had (and because I said I wouldn't do that and I just did :-( ) 15:44:46 with a force push you could rewrite history well into the past though, right? 15:44:52 so it depends on what you're rewriting 15:45:03 I could, and that would then cause you annoyance. 15:45:06 right 15:45:10 ok, i think i understand that better now 15:45:24 i'm not a git expert so my build script is probably not written in the ideal git way 15:46:09 Rebasing a branch anyone else has pulled is generally Bad and Wrong so the error here is mine, not in your build script. 15:46:44 well, i have no reason to maintain a local repository, i could easily just blow it away and pull your repo every time i need to build 15:46:54 that would mean no force pushing would ever be a problem on my end 15:47:01 That's vexingly slow though 15:47:07 i'm thinking that's probably what i should do since i'm not doing any local patches 15:47:15 yeah probably 16:24:35 Pinkbeast: made a pastebin account to get around that limit: https://pastebin.com/mUCVCYS7 16:25:24 Seems to be having some trouble loading ncursesw, but doesn't dgamelaunch use that? 16:25:39 I suppose not, those menus are just text. 16:27:54 libncurses6/testing,now 6.1+20181013-2 amd64 [installed,automatic], I think it's just trying different locations though. I don't see the open call that succeeds however. 16:29:23 pwnmonkey: Did you get to interact with it after running dgamelaunch? Did it display the menu? 16:29:56 Also check the libnucursesw5 package installed in the chroot 16:30:26 dgamelaunch works fine, once I'm too the crawl-git menu and try to (P)lay Crawl it just kicks me back to that same menu. 16:30:29 Ah good point. 16:32:28 For chroot: libncursesw5/stable,now 6.0+20161126-1+deb9u2 amd64 [installed] 16:34:17 Something else that might be useful: editing options with virus or otherwise has the same effect. 16:35:24 Can you in the chroot as user crawl run: 16:36:23 /usr/games/crawl-git-aaaaaaaaaa -name pwnmonkey -rc /dgldir/rcfiles/crawl-git/pwnmonkey.rc -macro /dgldir/rcfiles/crawl-git/pwnmonkey.macro -morgue /dgldir/morgue/pwnmonkey 16:36:37 where aaaaaaaaaa should be replaced with whatever actually exists there 16:37:09 /usr/games/crawl-git-9fb9971df1: error while loading shared libraries: libncursesw.so.6: cannot open shared object file: No such file or directory 16:38:13 Are you suuuure it's installed in the chroot? Humour me and "sudo chroot /home/crawl/DGL dpkg -s libncursesw5" 16:39:28 Package: libncursesw5 16:39:28 Status: install ok installed 16:39:55 Could you strace that crawl-git-aaaa invocation and pastebin the results? 16:40:25 Also tell me the Version line from the dpkg -s invocation 16:41:05 Version: 6.0+20161126-1+deb9u2 16:45:17 strace for crawl-git: https://pastebin.com/Hpk7eay4 16:47:17 What I find interesting here is that it's looking for libncursesw.so.6 but what you've actually got is a version of libncursesw5, as in Debian Stable, that gives you libncursesw.so.5 16:47:23 Right 16:47:37 Do "locate libncursesw.so" ? 16:48:03 blank under chroot 16:48:34 Sorry, I mean, run it not in the chroot 16:48:38 Will crawl use 6 during compile time if I only have 6 installed in the base system? 16:49:00 That is indeed the hypothesis I am forming. It's _built_ in the base system, but _runs_ in the chroot. 16:49:37 base system also returns nothing for `locate` 16:50:15 Well, "ls /lib/x86_64-linux-gnu/libncursesw.so*", where you might need to replace x86_64 with some other arch 16:50:30 Or do the dpkg -s outside the chroot and tell us the Version: 16:50:39 /lib/x86_64-linux-gnu/libncursesw.so.5 16:50:46 WTF 16:50:51 sorry 16:50:52 that was first line 16:51:01 5.9, 6 and 6.1 are there as well 16:51:20 Try and uninstall 6 and recompile? 16:51:31 does -dev come with 5 16:51:50 nvm I see it with 5 16:52:18 Are the base system and chroot the same version of Debian? 16:53:08 They are both Debian 4.9.130-2 according to uname 16:53:53 So you have a /lib/x86_64-linux-gnu/libncursesw.so.6 ? 16:54:04 If so, "dpkg -S /lib/x86_64-linux-gnu/libncursesw.so.6" 16:54:19 Under chroot? no. Only 5 and 5.9 16:54:35 Please only do things under the chroot if I ask for you to do so. 16:54:55 Yeah was just making sure. 16:55:14 So what's the answer? 16:55:49 libncursesw6:amd64: /lib/x86_64-linux-gnu/libncursesw.so.6 16:56:24 So I suggest either removing packages libncursesw6 and libncursesw6-dev from the base system, or installing them in the chroot, then recompiling Crawl 16:58:08 libncursesw6 doesn't seem to exist to apt under chroot 16:58:44 Well, you may want to just remove it from the base system, or look at discrepancies in /etc/apt/sources.list to work out where the base system got it from. 17:00:15 Do you think it would be enough to just remove *-dev? I have some dconf dependency on 6 in base. 17:00:23 Probably yes 17:00:58 Oh there isn't a -dev.. 17:01:11 <|amethyst> btw, uname doesn't tell you the debian version 17:01:30 Oooh yes, compare /etc/debian_version 17:01:33 <|amethyst> that's telling you the kernel version, which will be the same regardless of whether you're chrooted 17:02:22 base: buster/sid chroot: 9.9 17:02:31 Ahhh 17:03:54 I really think you want the chroot to be a like version to the base system. The easiest way to do this is probably to copy the /etc/apt/sources.list from base to chroot. Then you'll be able to install the libncursesw6 and -dev packages in the chroot. 17:03:57 so, debootstrap stable gives 9.9? 17:04:37 9.9 is the current stable. 17:04:55 buster/sid is testing/unstable and I dunno what it means when you've got both 17:05:08 Interesting 17:05:27 Doing the sources thing now. Thanks again for all the help. 17:05:49 It means "testing" (and idiot me, if you're in a room with the author of dpkg, you can just _ask_ these things) 17:16:17 Under dgamelaunch: https://imgur.com/a/ys9j8p0 :) 17:16:49 Let me reiterate that I presently can't do anything with Javascript, alas 17:17:16 Ah, it's just a picture of the species select screen. 17:17:21 Oh good. 17:17:40 Thank you very much for all the help. 17:19:28 No worries 17:20:41 I guess one more question though, does the ssh user that has dgamelaunch for shell need to be the crawl user? or can it be anything? 17:20:54 It can be anything - it's "fortunato" here, "terminal" on CKO, "joshua" on CAO etc 17:21:11 <|amethyst> it needs to be the same user that owns the save directories, etc 17:21:23 <|amethyst> but the actual name doesn't matter 17:21:56 can it safely just be the same crawl user? 17:22:02 |amethyst: Uh... the save directories here are owned by "crawl"; the user with dgamelaunch for shell is not "crawl". 17:22:23 pwnmonkey, things won't break but there are potential security concerns 17:22:54 <|amethyst> oh right 17:23:04 <|amethyst> forgot that dgamelaunch is chuid 17:23:53 <|amethyst> yeah, then the login name doesn't matter as long as dgamelaunch.conf has shed_uid = 17:24:30 <|amethyst> or shed_user = 17:33:39 Do you know if there is anything during the install process that removes /usr/sbin from path? That seemed to have happened during the course of install for me. 17:34:43 I do not know of any such thing; it is common for ordinary users not to have it in their PATH 17:56:22 and if dgamelaunch is the user's shell, it would need to add it to $PATH somehow because bash/etc. dotfiles won't be run 18:45:19 advil: i noticed +FB_BRAND didn't get documented 18:45:35 in 2153f8d6ca27ffb039a4da6617280d33973f9207 20:09:19 !lg 20:09:21 880. pwnmonkey the Martial Artist (L11 OpTm of Gozag), annihilated by a spiny frog on Lair:3 (st_stairs_11) on 2019-05-05 06:10:14, with 8952 points after 14410 turns and 0:29:16. 20:09:29 !help 20:09:30 !help: !help [CMD] displays help on a command. For a list of commands, use `!cmdinfo` for built-ins, and `!cmd -ls` for user-defined commands. 20:09:33 !cmdinfo 20:09:34 !apt !cdefine !cmd !cmdinfo !fn !ftw !function !gamesby !help !kw !learn !lg !lm !messages !nick !rng !seen !source !streak !tell !time !vault !won !wtf .echo 20:09:40 !lm pwnmonkey 20:09:42 6251. [2019-05-05 05:52:00] pwnmonkey the Martial Artist (L10 OpTm of Gozag) entered the Lair of Beasts on turn 12562. (D:9) 20:09:50 !source 20:09:50 Syntax is '[:[-]]', or the name of a function/#define/vault 20:11:33 troll with troll leather armour and ammy of regen 20:11:51 oops sorry, thought this was ##crawl 21:18:46 !seen gammafunk 21:18:46 I last saw gammafunk at Mon May 6 00:53:00 2019 UTC (25m 46s ago) quitting, saying 'Ping timeout: 252 seconds'. 21:59:58 -!- Taraiphest is now known as Taraiph