00:03:34 <minmay> nikheizen: usually spells above level 1 have multiple dice
00:03:53 <minmay> minor destruction is super variable because it can give you a level 1 spell or a level 3 spell
00:28:53 <nikheizen> minmay: hyperbole
01:28:56 <alexjurkiewicz> hellmonk: I wrote a 'differences between vanilla & hellcrawl' for a friend, it might be useful if you want to expand your intro in README.md http://dpaste.com/11ACXSR
01:29:14 <hellmonk> cool
01:29:16 <hellmonk> thank you
01:29:19 <alexjurkiewicz> I don't think it's complete, I made it by reading all your commits
01:30:28 -!- amalloy_ is now known as amalloy
03:31:17 -!- wHATEver is now known as Guest15762
03:40:02 -!- amalloy is now known as amalloy_
04:08:39 -!- Rjs- is now known as Rjs
09:53:36 <espais> ok so i temporarily fixed the 'out of pty' error i was getting the other day, but i don't know that its a permanent fix
09:54:03 <espais> i unmounted/remounted proc and dev/pts, then fixed the permissions on ptmx because for some reason it had gone to 777 from 666
09:55:32 <espais> is it possible that my cron job for recompiling the trunk overwrites some of the data from the mounting and ptmx?
10:21:34 <geekosaur> ptmx being 777 shouldn't have changed it, it's just pointlessly more permissive (adds execute permission)
10:22:06 <geekosaur> but it sounds like mount options are generally wrong, since that's where the permissions come from. check the mount options in the fstab?
10:22:33 <geekosaur> (this was also a problem last time someone was doing this on ubuntu if I read my irc logs from then correctly)
10:39:26 <espais> yea that's what i thought too, just thought i was odd that the perms had changed
10:40:01 <espais> so i followed the dgl guide for updating /etc/fstab
10:40:14 <espais> proc    /home/crawl/DGL/proc    proc    defaults        0       0
10:40:14 <espais> devpts  /home/crawl/DGL/dev/pts devpts  defaults        0       0
10:48:43 <espais> and if i look at the results hand-mounted
10:48:43 <espais> proc on /home/crawl/DGL/proc type proc (rw,relatime)
10:48:43 <espais> devpts on /home/crawl/DGL/dev/pts type devpts (rw,relatime,mode=600,ptmxmode=000)
10:51:13 <|amethyst> espais:  that's fine, the ptmxmode affects /dev/pts/ptmx, not the /dev/ptmx that programs actually use
10:51:32 <|amethyst> https://www.kernel.org/doc/Documentation/filesystems/devpts.txt
10:58:03 <geekosaur> that depends on whether /dev/ptmx is an actual device or a symlink
10:58:28 <geekosaur> iirc they have a symlink, my (based on an older ubuntu) has an actual device node
11:00:12 <|amethyst> espais:  hm, could you check?  is /chroot/dev/ptmx a symlink or a device node?
11:00:53 <|amethyst> espais:  if it is a symlink to pts/ptmx then you will need to change 'defaults' to 'ptmxmode=666' on the devpts entry
11:01:15 <|amethyst> espais:  (err, where /chroot = /home/crawl/DGL)
11:03:15 <geekosaur> 2017 Jul 18 17:46:37 <espais>	root@sbselab-git:/# ls -l /dev/ptmx lrwxrwxrwx 1 root root 8 Jul 14 15:58 /dev/ptmx -> pts/ptmx
11:03:20 <|amethyst> oh
11:03:39 <|amethyst> was that in the chroot?
11:03:48 <|amethyst> (though most likely they would have the same permissions)
11:06:22 <geekosaur> yes
11:12:08 <espais> fredericks@sbselab-git:~$ ls -la /home/crawl/DGL/dev/ptmx
11:12:08 <espais> lrwxrwxrwx 1 root root 8 Jul 14 11:58 /home/crawl/DGL/dev/ptmx -> pts/ptmx
11:20:30 <espais> |amethyst: i'll be back in about ~20 min if you have any suggestions
11:23:23 <|amethyst> !tell espais see what I said above; that's a symlink, so in the fstab entry for chroot/dev/pts you'll want to change 'default' to 'ptmxmode=666'
11:23:24 <Sequell> |amethyst: OK, I'll let espais know.
11:25:35 <|amethyst> !tell espais (also, you should ignore the permissions on the symlink itself, because they are almost always 777; chmod follows the symlink and changes permissions on the file it points to, which is why the chmod fixed things; it changed the 000 on chroot/dev/pts/ptmx to 666)
11:25:36 <Sequell> |amethyst: OK, I'll let espais know.
12:07:38 <espais> ok, back
12:07:38 <Sequell> espais: You have 2 messages. Use !messages to read them.
12:07:42 <espais> !messages
12:07:43 <Sequell> (1/2) |amethyst said (44m 19s ago): see what I said above; that's a symlink, so in the fstab entry for chroot/dev/pts you'll want to change 'default' to 'ptmxmode=666'
12:07:53 <espais> !messages
12:07:53 <Sequell> (1/1) |amethyst said (42m 17s ago): (also, you should ignore the permissions on the symlink itself, because they are almost always 777; chmod follows the symlink and changes permissions on the file it points to, which is why the chmod fixed things; it changed the 000 on chroot/dev/pts/ptmx to 666)
12:09:15 <espais> amethyst: so it should be like: devpts  /home/crawl/DGL/dev/pts devpts  ptmxmode=666    0       0
12:10:52 <espais> so i changed that, rebooted, still have the issue where it doesn't properly load
12:15:20 <espais> also very interesting
12:15:32 <espais> so i unmounted and remounted /dev/pts and /proc as before, still didn't work
12:15:41 <espais> changing permissions on ptmx back to 666 is what allowed the game to load
12:15:52 <espais> (which is kind of bizarre to me)
12:17:52 <|amethyst> hm
12:19:14 <espais> i stepwise checked, so basically i did the remount, refreshed the browser, nothing, then did the permission fix, refreshed, poof
12:19:27 <|amethyst> espais:  that fstab entry looks right:  when you do that, before chmodding, are the permissions on /home/crawl/DGL/dev/pts/ptmx ?
12:19:33 <|amethyst> s/are/what are/
12:20:14 <espais> heh
12:20:20 <espais> you forgot this:  %
12:20:30 <|amethyst> ?
12:20:44 <espais> vim command to kick off the search regex
12:20:52 <|amethyst> ah, I'm using sed :)
12:21:06 <espais> gotcha
12:21:07 <|amethyst> anyway, % means "apply to all lines", I just wanted to do the current line
12:21:16 <|amethyst> so really I forgot the :
12:21:22 <|amethyst> :)
12:21:27 <espais> ok so fresh reboot
12:21:50 <espais> c--------- 1 root root 5, 2 Jul 20 12:20 /home/crawl/DGL/dev/pts/ptmx
12:22:00 <|amethyst> hfm
12:22:03 <|amethyst> s/f/r/
12:22:16 <|amethyst> and the line in fstab is exactly as you pasted?
12:22:29 <espais> now it is
12:22:30 <espais> proc    /home/crawl/DGL/proc    proc    defaults        0       0 devpts  /home/crawl/DGL/dev/pts devpts  ptmxmode=666    0       0
12:22:46 <|amethyst> I hope there's a line break in between those two
12:22:51 <espais> yea
12:23:03 <espais> in my office now so i have to use the webchat..not necessarily sure how to get a line break on here
12:23:06 <|amethyst> that looks fine, wonder why it's being ignored
12:23:08 <espais> this is not a \n line break
12:23:10 <espais> yep
12:23:30 <|amethyst> does IRC even allow newlines?
12:23:34 <|amethyst> a
12:23:47 <espais> when i use pidgin i can get newlines
12:23:54 <|amethyst> ah, IRC doesn't allow CR-LF in messages
12:23:55 <espais> fredericks@sbselab-git:~$ mount | grep DGL
12:24:03 <espais> devpts on /home/crawl/DGL/dev/pts type devpts (rw,relatime,mode=600,ptmxmode=666)
12:24:04 <espais> proc on /home/crawl/DGL/proc type proc (rw,relatime)
12:24:17 <|amethyst> hmm
12:24:24 <|amethyst> that rules out one of my theories
12:24:38 <|amethyst> which is that it was getting mounted somewhere else
12:24:38 <espais> yea
12:24:42 <|amethyst> but it has the ptmxmode=666 line, so why is that being ignored?
12:24:56 <|amethyst> espais:  of, try  0666
12:25:02 <|amethyst> espais: s/of/oh/
12:25:04 <advil> I see the newline with my client
12:25:14 <|amethyst> advil:  the newline where?
12:25:21 <espais> do you see my password when i type it in?
12:25:22 <espais> hunter2
12:25:41 <|amethyst> huh, didn't show up.  Let me try:  bensonandhedges
12:25:48 <advil> in the thing espias posted, "proc..."
12:25:55 <advil> *pasted
12:25:56 <|amethyst> advil:  oh, huh
12:26:34 <|amethyst> advil:  I just see a space between the 0 and the devpts
12:26:39 <advil> test
12:26:39 <advil> test
12:26:48 <|amethyst> that was two messages
12:26:59 <advil> yeah, just testing how my client sends multiline messages
12:27:00 <espais> advil: you using webchat or a client?
12:27:08 <|amethyst> I guess webchat is willing to violate the RFC?
12:27:15 <advil> as is textual?
12:27:35 <|amethyst> advil:  you didn't send a line break, though
12:27:39 <espais> amethyst: changed to 0666 in fstab and no change to the permissions on ptmx
12:27:47 <advil> ah true
12:27:57 <|amethyst> advil:  Postel's principle
12:28:12 <|amethyst> whereas irssi seems to accept but not display the broken message
12:28:21 <|amethyst> or, it displays it, but not with a line break
12:28:30 <advil> same with the log bot, just has a space
12:28:37 <advil> anyways
12:28:43 <|amethyst> which I guess is good if someone writes a\nb\nc\nd\ne\nf etc
12:28:53 <|amethyst> since that wouldn't be caught by most flood protection stuff
12:29:08 <advil> hmm, wonder what textual would do
12:29:11 <advil> didn't mean to distract from espais' actual question :)
12:29:19 <espais> na
12:29:23 <espais> you're asking the real questions here
12:29:50 <espais> private DCSS servers for failover/load balancing testing are not relevant to dev chat ;)
12:29:50 <|amethyst> I'm kind of stuck, not sure what to tell you other than add a line to /etc/rc.local to chmod it :/
12:29:58 <|amethyst> err
12:29:58 <espais> yea
12:30:04 <espais> .profile?
12:30:06 <|amethyst> do remember when you change that you have to unmount and mount
12:30:11 <|amethyst> no, .profile runs on login
12:30:14 <espais> right right
12:30:15 <|amethyst> you want this to happen on boot
12:30:18 <espais> i want bootup
12:30:19 <espais> sorry
12:30:39 <espais> so this issue happened last time with an ubuntu servere?
12:30:41 <|amethyst> (and it has to run as root)
12:30:47 <espais> %s/servere/server/g
12:30:58 <|amethyst> espais:  no, but it's an older version of ubuntu where /dev/ptmx is not a link to /dev/pts/ptmx
12:31:08 <|amethyst> so the permissions on /dev/pts/ptmx never mattered
12:32:31 <espais> ah
12:32:38 <espais> i wonder if i should add this to the dgl wiki
12:33:04 <|amethyst> yes
12:33:06 <geekosaur> I am wondering if there's a devfs rule or w/e overriding the mount option
12:33:12 <|amethyst> that's possible
12:33:21 <|amethyst> it would be nice to figure out the "correct" way to do it
12:34:28 <|amethyst> what does the mount entry for the host /dev/pts look like, btw?
12:34:35 <espais> i'm willing to be a test case
12:34:41 <espais> in fstab?
12:35:01 <|amethyst> no, in the output of 'mount', on in mtab
12:35:11 <|amethyst> well, I guess the entry in fstab would be good to know too
12:35:26 <geekosaur> might be interesting: grep -r ptmx /etc/udev
12:35:28 <espais> ah, sent that prev...will do again
12:35:45 <espais> geekosaur: no result
12:35:56 <espais> fredericks@sbselab-git:~$ mount | grep DGL devpts on /home/crawl/DGL/dev/pts type devpts (rw,relatime,mode=600,ptmxmode=666) proc on /home/crawl/DGL/proc type proc (rw,relatime)
12:36:01 <geekosaur> ok, so udev is *probably* not messing with it
12:36:23 <|amethyst> espais:  no, the one on the host
12:36:25 <|amethyst> espais:  /dev/pts
12:36:30 <espais> fredericks@sbselab-git:~$ mount | grep pts
12:36:35 <espais> devpts on /dev/pts type devpts (rw,nosuid,noexec,relatime,mode=600,ptmxmode=666)
12:36:37 <espais> devpts on /home/crawl/DGL/dev/pts type devpts (rw,relatime,mode=600,ptmxmode=666)
12:37:09 <|amethyst> well, nosuid noexec wouldn't hurt to add, but that doesn't explain why they come out different
12:41:22 <espais> hm
12:44:04 -!- amalloy_ is now known as amalloy
12:44:49 <geekosaur> unless something is being 'clever'
12:45:02 <espais> that's what i'm wondering
12:46:07 <geekosaur> ("something" including but not limited to systemd, which slipped into ubuntu some time ago --- I don't know how integrated it is in recent ubuntu though. debian's been mostly keeping it on a tight leash...)
12:49:55 <espais> so systemd has a pretty heavy hand on how webtiles is kicked off...
12:52:30 <geekosaur> I;'d be very surprised if it were remounting devpts though; its style is more using cgroups to limit the permissions visible to things
12:52:34 <espais> ugh, i can't seem to register for the wiki
12:52:37 <espais> what chars would you say this is
12:52:42 <espais> http://imgur.com/a/U0ag6
12:53:00 <amalloy> R E K T
12:53:24 <amalloy> but i dunno, if i had to guess for real, it's BfCff
12:53:31 <geekosaur> B F C F F
12:53:38 <geekosaur> yeh
12:53:46 <geekosaur> that's a crappy captcha though
12:54:12 <espais> it doesn't seem to think that those are F's
12:54:31 <espais> keep getting confirmation hash doesn't match
12:55:50 <geekosaur> that or the thing we're seeing as a C is a confounder instead of a character
12:55:58 <geekosaur> since its top is overlapped
12:58:46 <espais> very good at keeping out humans and robots alike
12:58:57 <|amethyst> does the wiki have its own accounts?  I figured it shared with mantis
13:01:07 <espais> well i got in...refreshed until that thing wasn't in the security string
13:05:49 <amalloy> |amethyst: a thing i hear every so often on reddit is that people want to edit the wiki but can't make an account without emailing an administrator, who apparently doesn't read email very often
13:06:13 <|amethyst> amalloy:  that's the other wiki
13:06:23 <|amethyst> though I have seen this Bfff thing before I think
13:20:41 <johnstein> more like craptcha
13:47:26 <espais> ^-- the real hero here
13:55:48 <espais> so i updated my rc.local file with some of those changes and now it works like a champ
13:55:54 <espais> i edited the wiki appropriately
13:55:59 <espais> hopefully that helps somebody in the future
14:35:04 <Lightli> ??c-r-d
14:35:04 <Sequell> c-r-d[1/2]: https://lists.sourceforge.net/lists/listinfo/crawl-ref-discuss
14:48:15 -!- Taraiph is now known as Saraiph
14:58:13 -!- Saraiph is now known as Taraiph
15:00:35 <espais> hey all, quick thought-experiment
15:00:46 <espais> if i wanted to do some load balancing / failover with a webtiles server, do you think it would break anything if i did a shared filesystem type approach?
15:00:59 <espais> where i basically mirror /home/crawl and /home/crawl-dev across multiple servers?
15:01:16 <espais> i've done that for some basic website testing, but nothing involving the filesystem and pids like crawl does
15:41:08 <|amethyst> it would need to be something like nfs or lustre or gfs that supports it
15:41:27 <|amethyst> err
15:41:35 <|amethyst> that supports shared locks
15:42:10 <|amethyst> and there's other stuff that's not shared through the filesystem
15:42:33 <|amethyst> the unix domain sockets for example between webtiles and crawl wouldn't work over the network
15:42:37 <|amethyst> you'd have to change that
15:42:59 <|amethyst> or do something to ensure the different server's processes don't step one one another
15:43:29 <|amethyst> e.g. move the sockets under /var/run instead of the chroot's webtiles directory
15:44:12 <|amethyst> probably everything related to webtiles, in fact
15:45:22 <|amethyst> if you're using a shared chroot, you probably need to mount or bind-mount the hosts' /var/run beneath the chroot's /var/run
15:45:37 <|amethyst> well, some host-specific thing
15:45:51 <|amethyst> it probably shouldn't be the host's /var/run because that's a potential information leak
15:46:32 <|amethyst> dgamelaunch user db and version db would also need to move to such a per-host filesystem
15:46:39 <|amethyst> hmm
15:47:02 <|amethyst> or maybe version db should be shared, but I don't know whether sqlite3 can handle that
15:47:22 <|amethyst> I guess same goes for user db
15:48:05 <|amethyst> dgamelaunch and webtiles config do need to be per-host unless you're going to listen on 0.0.0.0
15:48:29 <|amethyst> especially if you're using https
15:49:04 <|amethyst> oh, you said load-balancing, so they can all have the same hostname and https is fine
16:04:29 -!- aakz is now known as aak
16:05:21 -!- aak is now known as aakz
17:35:07 <Raine357> is it possible to attach a lua debugger to crawl in order to do things like set watch on lua variables
17:35:21 <Raine357> or would it have to be done with a C++ one
17:39:17 <|amethyst> would be nice to have, but we don't have it
17:39:42 <|amethyst> it's not really fun or easy to debug lua from the C side
17:40:31 <|amethyst> I guess it's less terrible than some languages would be
17:42:09 <Raine357> ok
17:43:47 -!- amalloy is now known as amalloy_
17:55:03 -!- amalloy_ is now known as amalloy
18:15:25 -!- Menche_ is now known as Menche
20:21:29 -!- amalloy is now known as amalloy_
20:39:22 -!- amalloy_ is now known as amalloy
21:22:08 -!- GiantOwl is now known as Kalire
21:22:10 -!- Kalire is now known as Kalir
21:52:02 -!- amalloy is now known as amalloy_