01:24:44 Unstable branch on crawl.develz.org updated to: 0.22-a0-362-g535178c (34) 01:57:24 Windows builds of master branch on crawl.develz.org updated to: 0.22-a0-362-g535178c 02:53:44 Monster database of master branch on crawl.develz.org updated to: 0.22-a0-362-g535178c 03:03:39 03alexjurkiewicz02 07https://github.com/crawl/crawl/pull/721 * 0.22-a0-354-g27845c7: species: Remove species_order 10(4 minutes ago, 1 file, 33+ 61-) 13https://github.com/crawl/crawl/commit/27845c7e4680 04:36:18 -!- amalloy is now known as amalloy_ 10:36:46 Hey guys if I want to change the port on webtiles, is there someplace I need to change it other than config.py? 10:37:32 if you're using a reverse proxy setup, you need to change that 10:38:06 (e.g. nginx or apache sitting in front of crawl, to provide an info page without needing to use a vhost) 10:38:14 yeah I did that but the server isn't switching it's still trying to launch on 8080 10:46:59 if you're running it in place from the crawl build directory, config.py is definitely enough (just tested it) 10:47:52 I don't recall if dgl config might be involved 10:49:53 where would I find that, I am using DGL 10:50:45 tbh the first thing I would do is probably something like "grep 8080 */*" or w/e 10:50:59 not knowing much about DGL setup 10:51:22 MikeHollis: are you saying you're using the dgl based setup we provide? 10:51:27 dgl itself has nothing to do with the webtiles port 10:52:02 https://github.com/neilmoore/dgamelaunch-config/blob/szorg/config.py 10:52:46 yes gamma 10:52:46 if you are using our dgl setup scripts to manage your webtiles installation, you need to modify this file in your checkout of the dgameluanch-config repo and then publish 10:52:52 so that the templated file gets substituted with the right data and gets installed in the chroot 10:53:29 https://crawl.develz.org/wiki/doku.php?id=dcss:server:setting_up_dgamelaunch_and_webtiles see the info starting on step 8 10:53:50 and step 10 gives you the command to publish the template into the chroot 10:54:02 so in general I recommend you reference those instructions when trying to modify something 10:54:47 That document is sadly a bit incomplete, although like an idiot I didn't write down the extra changes I had to make :-( 10:55:08 the instructions relevant to this discussion are fine 10:55:41 there are simply a lot of steps involved and the details of the OS and the desired configuration can be different from system to system 10:57:07 MikeHollis: it sounds like you might have modified the template version and not published it; publishing the template installs the final version of config.py into the chroot 10:57:20 not only this file but other files as well 11:01:43 this doesn't make any sense 11:01:44 hang on 11:03:03 https://pastebin.com/XR8GCD2q 11:06:35 MikeHollis: the way you did that log it almost looks lik you didn't restart webtiles 11:06:38 did you restart it? 11:06:58 that is, are those new log entries? 11:07:08 what I would verify is that and check the config file in the chroot 11:07:14 was it actually updated to have the new port? 11:08:02 yeah I did restart it I forgot to include that part int he pastebin, I also just tried it again to make sure and it still is on 8080 11:08:47 MikeHollis: are you using ssl? is the ssl port binding updated if so? 11:08:52 I'm not using it, so I havn't touched it, 11:09:03 and I guess my earlier question about whether the final chroot config.py was properly updated still applies 11:09:11 do you see the new port in the installed config.py 11:09:20 checking one sec 11:09:43 hrm, there's bind_port and bind_pairs 11:09:53 I wonder if bind_pairs is overriding 11:10:08 hmm nope that's the failing point 11:10:13 ah, ok 11:10:17 bind_pairs is commented out by default, I think 11:10:23 it's not installing to the published dir 11:10:27 I should really set one of these up myself some day 11:10:29 advil: not in the szorb branch 11:10:33 ah 11:10:34 of that repo 11:10:45 but yeah it might be in MikeHollis' version 11:10:51 sounds like his config.py is not getting updated 11:11:01 s/szorb/szorg/ 11:11:02 yeah that's exactly what's happening 11:12:11 what's the location that sets what gets copied over? I forget where it's at, trying to fix an old installation 11:12:17 in your pastebin there are two places where a config.py is getting copied 11:13:18 dgl-manage.conf I think 11:13:31 actually hold on it might be simpler than that, 11:13:42 https://github.com/neilmoore/dgamelaunch-config/blob/szorg/dgl-manage.conf#L16 11:17:01 okay I was editing in the src, and not in dgamelaunch-config/config.py 11:18:22 that was it 11:18:45 everytime I make adjustments I forget what files come from where, the port updated that time 11:19:46 bleh, still failing websockets connection though even on a new port 11:19:52 yeah that's a frequent source of confusion I think 11:20:24 can I safely symlink those two files? 11:20:29 the config has to be templated so the template is kept out of the source 11:20:42 no, you mean the webserver source file? 11:20:50 hrm, well maybe you could I guess 11:21:05 since this file doesn't get copied (the config.py in the src, I mean) 11:21:12 I'm not actually sure how that bit is handled 11:21:23 fair enough 11:21:33 whatever you do, the config.py in that directory in the repo must be a template 11:21:41 and not the version in the source 11:21:54 I think you'd not want to do what you said though 11:22:01 you'd be modifying the source in the crawl repo 11:22:23 so when you updated you'd get conflicted from unstaged changes I think? 11:22:36 anyhow, yeah you might want to make a note about it in your notes for setting this stuff up 11:25:17 if it's not used, and I create a patch does it really matter? 11:26:58 so here's another question I have, the log says the server is started on 8180, but I am not making websocket connections and when I try to pull up the port in a browser directly, it acts like the server isn't running 11:27:20 If you locally patch it every time it should be fine, yeah, just saying that you need to do something otherwise you'll get a conflict when the source updates for the crawl repo 11:27:28 this is theh same issue I had before the port change and why I changed ports so it's very likely a different issue 11:27:50 gotcha, and yeah I've got that part covered already 11:27:55 sounds like a configuration issue with your firewall, yeah 11:28:13 you can try from the machine itself and verify that it's listening on your port 11:28:23 ugh! you're right, stupid iredmail, I keep forgetting it cleared my firewall settings 11:29:11 and boom just like that I'm fixed 11:29:14 lol 11:29:20 \o/ 11:29:28 thank you so much 11:29:31 np 11:30:05 no loss on the port change, 8080 is too popular anyway 11:33:00 I guess the script could check that other file and print a warning if it's been changed from master, if this is a common issue 11:34:28 advil: I think it would be better to just update the notes in the guide 11:34:51 not that the scripts couldn't use a lot of work but that would have it examining the crawl repo which is not a thing it currently does 11:35:27 from observing these discussions I uh suspect we might have reached the point of diminishing returns for just adding stuff to the guide :-) 11:35:37 it's not really that simple advil 11:35:48 you have to really understand how these scripts work first I think 11:36:06 what I would advise is to go through the installation yourself and understand those scripts before attempting modifications 11:36:20 well, I'm not particularly planning on modifying anything right now 11:36:20 they're kind of non-trivial with a few different relevant scripts involved 11:36:43 but I think just a note in that section of the guide warning that config.py in the repo is not the same as the one in the source would be helpful 11:37:09 and I'm probably overstating how common this was as a problem, I think it's happened once before 11:37:33 just it's confusing to people, the entire way this is set up, as to what is a template and what is not 11:37:52 also the szorg branch is the branch people have to use since it's the most current, while master is actually not 11:38:17 but I haven't tried to merge them; |amethyst was the last person to make extensive modification to the scripts 11:38:35 or maybe by extensive I should say "any modification at all" 11:39:08 I only know them sort of well because I made a branch to make them work with webtiles-changes back when that branch was trying to be a thing 11:39:57 that setup is pretty confusing, and some people have made some attempts to use other technologies to manage the setup 11:40:04 yeah, I've seen the ansible thing 11:40:46 I do plan on setting up a server at some point 11:40:46 but it looks painful 11:41:17 it's not actually *too* painful but the setup is just very all over the place 11:41:23 yeah 11:41:58 the scripts and what they do are pretty cool, even if they don't cover everything, but yeah some sort of more modern ansible solution or other kind of similar solution might be really helpful 11:42:13 the last person to do extensive work on that kind of did it by themselves and left without trying to get an admin to use it 11:42:25 my plan is to start out by trying to do it in a container so I can dodge some of the more random idiosyncratic setup problems people seem to hit 11:42:33 which is sort of what we need, for official servers to actually use the new setup 11:42:38 yes 11:42:41 I agree 11:42:44 I think my major issue with it was that I had no idea how everything related to each other, so when I was going through the setup process initially, I had a lot of questions, 11:42:50 though I fully understand the lack of volunteers 11:43:01 it was like putting together a puzzle with a box image that was scrambled 11:43:07 heh, yeah 11:43:11 heh 11:43:50 the scripts are pretty nice and do work well once set up and when you understand them, but it's very confusing initially 11:43:56 in fact for the pghosts thing I really wish I already had this because the blocker is trying this out on a server and making sure it doesn't go boom (without wasting an admin's time) 11:44:14 oh right, I need to do some testing, that's what I wanted to work on today 11:44:21 I was thinking I could up ghost generation 11:44:32 and just try out a game and see what happens 11:44:49 perhaps I could even do a full dgl setup and try it 11:44:51 ah...I set it pretty low (because in practice the current rate just provides an upper bound) 11:45:03 yeah, but if I wanted to say always see a ghost vault on a level 11:45:09 I can just chance that 10% to 100%? 11:45:18 yeah 11:45:21 to see the permaghost thing in action 11:46:06 you'll need to die some, if there's no ghosts at all anywhere it'll fall back on mirroring the player 11:46:06 ok gotta run, late for a meeting, later 12:00:49 what's the path for websocket connections? 12:01:03 url I mean 12:02:34 I am not a fan of having a visible port so I'm proxy'ing but the websocket connections arn't making ti through the proxy, I'm pretty sure I can just add that to the virtualhost but I don't know what path they look for 12:03:04 well there's a websocket url syntax 12:03:13 but I'm not sure your proxy configuration wants to see that 12:03:23 it's just over http (or https if you're using that) 12:03:30 but yeah I guess there is a specific url 12:03:45 var socket_server = "ws://crawl.berotato.org:8080/socket"; 12:03:47 for cbro 12:03:58 shows up in the javascript of the lobby html 12:04:06 so you could replace that hostname/port with your info 12:04:19 that might be what I'm looking for 12:04:20 ty 12:04:21 hence http://HOST:PORT/socket 12:12:21 well that's not working on public or localhost interfaces, but it does work directly so it's on me now, I appreciate the help getting eveything going again 12:35:43 huh, kind of interesting result of the rations change that I hadn't thought about, it's possible to use all your permafood with fed and get in trouble if you aren't paying attention (https://www.reddit.com/r/dcss/comments/8dk9oq/i_might_starve/) 12:36:05 (the player did make it) 12:46:14 <|amethyst> sounds like a good thing to me :) 12:49:30 yeah, not saying it's bad 12:50:13 this sounds like a good result for sure: "Just went for the hell runes. You wouldnt believe how i felt when seeing any corpse leaving monster. Each chunk was like finding the most epic unrand. Holding chunks in the inventory made me feel like a god. " 13:35:48 -!- amalloy_ is now known as amalloy 15:36:54 -!- pikaro_ is now known as pikaro 17:40:20 is there a preset way to alert a user to unvisited upstairs? if not: there are some vaults that can only be see through upstairs, and the interface already notes what upstairs connects to which downstairs and if they are known. it's therefore possible but tedious to find out if you missed any of them. as per the games design strategy, there should probably be a way to find unexplored upstairs on levels below seemingly fully 17:40:20 explored ones. is this worth a feature request? 17:57:29 pikaro: they are marked as unvisited 17:58:22 advil, that's what I refered to in the rationale. but is there a way to display that in some way for the whole dungeon? 17:59:25 across the whole dungeon, I don't think so 18:00:27 you can use X to quickly check, I gues 18:00:28 s 18:00:29 hence the "tedious" part - you can easily go through the whole dungeon and look at every stair without much of an expense since there's no real food clock 18:01:43 well you can use X to do this without moving 18:02:34 ah I'm used to webtiles, that takes a really long time there. forgot there's also an offline version :P 18:03:25 ~5s per level with all the branches and periodically checking to make sure adds up 18:03:42 ah, webtiles X doesn't show the stair symbols 18:03:49 (console does) 18:04:34 in fact webtiles X doesn't even show the level name 18:04:34 oh offline does more? I was thinking go <<<< on every level, see if something is marked * 18:05:05 plus loading times of course 18:05:12 yes, console (online and offline) already shows the stairs symbols for the level, so you can count and see visited, not sure what offline tiles does 18:09:00 alright I wanted to clone offline tiles anyway to get familiar with the new code, I'll find out how they differ and maybe submit a feature request accordingly. thanks! 20:01:06 -!- amalloy is now known as amalloy_ 20:25:56 -!- amalloy_ is now known as amalloy