00:17:17 is there a place to modify the html template files that plays nice with dgl publish 00:18:18 -!- Tux[Qyou] is now known as ishaplzban 00:18:19 or are they safe from updates 00:18:23 -!- ishaplzban is now known as Tux[Qyou] 00:44:09 Does anyone know how to turn off this message, I'm only playing on one branch, and I am using dgl: Trunk games don't count for the tournament, you want 0.21. Play trunk anyway? (Y/N) 00:45:01 I just want that to always assume Y 00:46:33 Huene`: you'll need to revert that command 00:46:51 *commit 00:47:09 I don't think there's a great facility for those dgl scripts to apply local patches 00:59:48 -!- rchandra1 is now known as rchandra 01:01:53 how often does the makefile change? 01:02:56 because I can correct the tourney thing in the makefile that's how I've fixed it in the past, but with the directories changing all the time I need to do something different 01:03:41 essentially all I need to do is unset TOURNEY 01:05:26 Huene`: you can answer that question for yourself by running `git log Makefile` 01:15:29 Huene`: makefile can change just randomly, so I wouldn't count on it not changing 01:15:50 what I guess you could do is have the dgl script look for a patch file 01:15:56 in some designated location 01:16:21 and then apply it if it exists, after the git pull 01:17:05 problem is I never understood how patch files worked, and after finally getting to sit down and play I think I'm done thinking for tonight, I also figured I could probably sed the makefile to make the neccesary change, I just need to do it after it downloads but before it compiles 01:17:28 https://github.com/neilmoore/dgamelaunch-config/blob/master/crawl-build/update-public-repository.sh#L19 01:17:59 oh cool thank you that helps 01:18:01 Huene`: I would look at simply modifying this script to test for the existence of something like local-changes.patch in the directory 01:18:15 if said file exists, git am local-changes.patch 01:18:18 after the pull 01:18:37 the problem is if there's a conflict, of course 01:18:44 ideally the script would detect this and do something sensible 01:19:05 probably just generate an error is fine 01:19:25 I don't mind it occasionally messing up, but I have it updating daily, so I don't want to manually change things daily 01:19:56 yes, so changing this function would make it so you don't have to do things manually 01:20:06 until the patch is no longer relevant, that is 01:20:10 Unstable branch on crawl.develz.org updated to: 0.22-a0-79-g605b837 (34) 01:20:19 when that happens though, you could delete the patch 01:20:29 and the script would not try to apply it 01:40:28 gammafunk: does the script actually run git pull? that will merge local changes anyway, with no need to change the script at all 01:40:39 no patch file, just make a commit to trunk 01:41:35 oh, you linked the script 01:42:19 it looks to me like update-crawl-ref would work perfectly fine in the presence of local commits. you'd just have to do something to resolve conflicts if and when those ever occur 01:53:02 amalloy: hrm, maybe you're right 01:53:15 just commit to the branch and git pull works normally 01:53:26 obv you can have conflicts but it's the same situation 01:53:41 Huene`: dunno if you saw what amalloy said, but that's probably the way to go 01:54:05 I think just having the script abort would be fine 01:54:09 it's running under set -e 01:54:25 so just git pull returning an error (which I think it does) would do the trick 01:55:59 well the problem is I'm not on a set version, I'm on the most up to date commits, so everytime it downloads one it does it to a new directory 01:56:15 but that's good to know 01:57:09 Huene`: no, that's not how it works 01:57:17 or generally not how it should work 01:57:23 it has a single clone of the repo 01:57:37 you have a master branch in that repo which is used to pull the latest version 01:57:40 that's the "downloading" 01:57:51 amalloy is saying that you make a commit in master 01:58:17 so when your git pull command proceeds to get the latest changes to master 01:58:23 it will merge them on top of your special commit 01:58:38 hm 01:58:41 well 01:59:18 so here I assume? /home/crawl-dev/dgamelaunch-config/crawl-build/crawl-git-repository/crawl-ref/source 01:59:18 oh, i see. yes, that's still true. i thought it was re-cloning each time 01:59:29 but it just re-clones if it has to 01:59:30 no it does not reclone each time 01:59:37 that is the initial clone, yeah 01:59:49 so that'd be where I'd want to make changes? 01:59:59 Huene`: that's the repo, yes 02:00:27 it has branches for every crawl version you have installed through the dgl scripts 02:00:35 Windows builds of master branch on crawl.develz.org updated to: 0.22-a0-79-g605b837 02:00:36 well it has those branches regardless of that, but 02:00:40 every time it does an update 02:00:58 it cds to that repo dir 02:00:58 checks out the relevant branch (master branch for trunk) 02:01:01 and does git pull 02:01:03 I'm just playing on the up to date branch 02:01:03 to get the changes 02:01:08 yes, that is trunk 02:01:14 which is on branch master 02:01:34 we call the master branch trunk, but the name of that branch in git is 'master' 02:01:41 so your commit is made in the master branch 02:02:06 note that git branches are not a specific directory, they are tracked by git 02:02:11 from the repo directory 02:02:22 and yeah, you'd revert that tournament warning commit in your master branch 02:02:35 just do: git revert HASH 02:02:41 where HASH is the commit that added it 02:02:46 !gitgrep 1 [Tt]ournament 02:02:47 %git HEAD^{/[Tt]ournament} 02:02:47 07wheals02 {GitHub} * 0.22-a0-44-geb883ae: 0.21 tournament reminder. 10(11 days ago, 1 file, 1+ 1-) 13https://github.com/crawl/crawl/commit/eb883ae61ce9 02:03:05 so you'd: git revert eb883ae 02:03:15 after having done: git checkout master 02:03:17 in that repo dir 02:03:44 this would add a new commit removing the reminder 02:04:05 I'd still get updates though aside from that correct? or does it roll me back to that time and leave me there 02:04:21 it adds a new commit, no it doesn't roll you back 02:04:31 I think you don't quite understand how git works, so this is a bit confusing to you 02:04:46 but changes come in through git to your master branch 02:04:51 yeah I have no idea how git works, I've tried to figure it out several times and it always gives me a headache 02:04:53 you are adding a change undoing a commit 02:05:13 and git pull is a command that pulls in new commits (in this case from our repo) 02:05:18 and merges them into your master 02:05:31 that's how you're getting these updates 02:05:40 the only twist is you now have this special commit you made 02:05:44 which removed the reminder 02:06:15 which is fine, git pull will be ok with that and apply the updates on top 02:06:20 eventually though we'll remove the reminder 02:06:44 and when that happens you'll want to do a command to reset the master branch to ours, which is easy to do 02:08:50 heh, well that's nice to know, my workaround undoes this in exactly the same this particular commit was adding it doing 02:09:22 yes, that's what the git revert I mentioned would do as well 02:09:39 it would remove exactly what that commit added 02:10:11 and you could just do the change yourself and make a commit of that change 02:10:20 git revert is just a shortcut to undoing a commit 02:10:34 well. i would try to avoid the words "undoing a commit" 02:11:03 sorry, reunaddifying backwords the anti-commit 02:11:09 thank you 02:11:33 git reunaddify is a command that should be more well known 02:11:42 I'm imagining the man page for it now 02:11:53 gammafunk: https://git-man-page-generator.lokaltog.net/ 02:12:09 yeah that's a great site 02:14:07 !messages 02:14:08 No messages for TZer0. 02:43:32 so one thing i've noticed during my recent love affair with SpEn is that a lot of uniques retain their weapons when tuk's dance is cast....is this related to the sprite itself or just missing code relating to the doll parts? 02:44:09 -!- Evablue__ is now known as Evablue 02:50:52 i think uniques often don't have doll parts. the sprite contains the equipment they usually start with 02:51:04 eg, https://github.com/crawl/crawl/blob/master/crawl-ref/source/rltiles/mon/unique/sigmund.png 02:51:28 yea that's what i was wondering 02:51:50 so fixing a lost weapon would involve probably fixing all the sprites as well 02:51:52 compare to https://raw.githubusercontent.com/crawl/crawl/master/crawl-ref/source/rltiles/mon/humanoids/humans/vault_sentinel.png which has space for at least a weapon 02:52:29 ok, thanks 02:53:03 FYI, is single sign-on between crawl servers something you're really interested in (re your RL talk you gave a little while back)? 02:53:04 yeah, and even if you did that you couldn't really achieve the current look for sigmund. remove his stylised scythe and add a doll slot, but that means he'll just have a standard scythe and not a cool one 02:53:58 i'm sure that the doll parts code could be extended to handle uniques though 02:54:24 Monster database of master branch on crawl.develz.org updated to: 0.22-a0-79-g605b837 02:54:44 espais: i think it's something we'd be happy to have. i don't know how much thinking has been done about what the ramifications of single sign on would be 02:54:52 !tell Huene` any luck with the dgl install after i took off? 02:54:52 espais: OK, I'll let huene` know. 02:54:56 nice. 7 panlords on my brother's zig 17 02:55:03 i was trying to puzzle that one out the other day 02:55:08 https://imgur.com/a/3Wm9W 02:55:15 oh. wrong channel. sorry 02:55:20 ?/webtiles-changes 02:55:20 Matching entries (2): nfm[1] | nfm[2] 02:55:26 ??nfm 02:55:26 nfm[1/7]: <+gammafunk> |amethyst: I've a patch for the webtiles-changes branch on dgamelaunch-config: http://sprunge.us/HONd?diff 02:55:31 ??nfm[2 02:55:31 nfm[2/7]: short-term server TODO: dgl-config setting for webtiles server branch; update stable and trunk scripts to match; switch to webtiles-changes; Chei -> dobrazupa; autobuild glyphs+monster; clean up explbr, add charms-reform faithful salamander no_backtracking_god 02:55:36 afaik, each crawl server uses an sqlite database to track users, so there'd have to be some sort of controlling database and new fields inserted into each separate database probably at minimum 02:55:49 i vaguely recall webtiles-changes being discussed in connection with SSO, but i don't think they're actually related 02:56:17 espais: well, or we could make the servers not use an sqlite database, and query whatever the central database is 02:56:43 i love the idea of sso but that introduces a whole range of separate issues, eg, what if the central server goes down 02:56:53 yes 02:57:10 (which might give strength to using a local db as a main login and just tie back to the central for unique id management or something along those lines) 02:58:17 plus there would have to be some sort of authorization for each and every person (like me) who decides to setup a private dgl server 02:58:21 perhaps, indeed. as you can see i'm no good at designing this 02:58:34 i may also be overthinking it too 02:58:47 but i did agree with your point, it would be nice to have a unique username across all public servers 02:59:15 and if i feel the need for some banging tunes, i should be able to easily login to the korean server :) 02:59:46 i'll puzzle it out for a bit and see if i can come up with something...like i said i have a local dgl install so i could use that as a testbed 03:00:56 anyway, i suppose i should get some sleep so i can get up for work in the morning, have a good one 03:10:58 I know there's a lot of issues with central auth, but as a server owner, I welcome the day that I don't have to worry about user auth and user management 03:11:40 Unstable branch on crawl.beRotato.org updated to: 0.22-a0-79-g605b837 (34) 03:14:04 johnstein: a nice thing about central auth is that it implies the existence of central anything. suggesting we could just have one server saving ttyrecs or whatever 03:14:24 yea. that whole system needs automated as well 03:15:12 after talking with |amethyst again lately and realizing that I don't need to be so careful on which ttyrecs I move over, I should just update my compress-ttyrec script to rsync files over to shaolott automatically 03:15:18 after they compress to bz2 03:16:11 I just need to know what the Right Way to do that is (i.e. how to ensure that all files got copied over before I delete locally. I really don't like deleting files automatically 03:16:13 ) 03:19:05 johnstein: you could use rsync, which looks like it returns 0 iff all operations were performed successfully 03:19:33 I figured there was an exit code I could use. just haven't actually looked into it 03:19:44 (I use rsync already) 03:19:50 ^status 03:19:50 25 Crawlers. CBRO disk usage=90% (135GB) | RAM usage=26% (4GB)| uptime/CPU= 03:19:48 up 10 days, 7:55, 1 user, load average: 0.57, 0.63, 0.63 (4 Cores) http://status.berotato.org 03:19:59 the next thing I need to do is force old versions to update 03:20:22 rsync --stats will print a lot of useful stats for you 03:20:22 but I need to know more how that works 03:20:22 ooh nice 03:20:22 ty 03:20:42 actually reading the man page more carefully i think s/useful/useless/ 03:20:53 heh 03:21:16 is there a way for me to force update player games to a certain version? 03:21:48 I think I remember someone in here suggesting that to me as a way to remove a lot of the old versions (clean-trunks script only removes versions if there are no current saved games for it) 03:23:19 i don't know how server admin goes. i imagine you can 03:34:21 johnstein: i'd say, if you're worried about relying on the output of rsync, try adding something to your script or whatever that checks the return of rsync, and tells you whether it *thinks* it would be safe to delete the local ttyrecs 03:34:49 run it that way with whatever manual human verification process you currently use, until you're satisfed that it's doing what you think 03:34:57 yea, that's my backup plan. that way the worst case is... yea, what you just said 03:35:08 should be easy enough to verify 03:35:46 my plan is to create some nice system diagrams of all the server admin knowledge I know, then give it to someone who actually knows what they are doing to suggest a more sane system 03:36:20 the secret is, that's how all system administration is done, and nobody actually knows what they are doing 03:37:12 like, I can't remember how I set up the cron job for the clean-trunks script 03:37:52 I know where the script is, and I know the log file it's making, but visudo (where I thought I set it up) is pointing to scripts at other locations 03:39:47 hmmm, looks like maybe the clean-trunks script hasn't been working properly since.... 03:40:19 nov 23, 2016 03:41:03 %git 173670c088 03:41:03 07reaverb02 * 0.15-a0-350-g173670c: Make blink frogs SIZE_MEDIUM 10(3 years, 9 months ago, 1 file, 1+ 1-) 13https://github.com/crawl/crawl/commit/173670c08884 03:41:13 says it can't find that revision 03:41:32 but it exists! 03:45:59 what is clean-trunks? 03:46:50 it's a dgl script that cleans out trunk builds that are no longer neaded 03:47:06 this is why cbro has no room: http://crawl.berotato.org/crawl/dev/trunk_versions.txt 03:47:25 dgl script meaning the set of scripts in that dgamelaunch-config repo to manage our dgl setup 03:47:46 (re: the error I saw, the binary was missing but the data existed in crawl-master, so it complained. touched a blank file with the binary version name and it cleaned up ok) 03:48:12 we just gotta bump major version... 03:48:20 oh I guess that doesn't actually address that 03:48:26 since those people just won't ever be able to transfer 03:48:39 and hence still need their build 03:48:58 can I upgrade them to the latest version right up to before it would break them? 03:49:11 I'm really weak on this aspect of compatibility and updating saves 03:49:12 you could, yes 03:49:23 I guess if you updated them all to a common version 03:49:25 so I'd need it spelled out pretty carefully to me 03:49:30 it would allow you to save space 03:49:44 yea, that's what I was hoping. I should be able to implement some policy on that 03:50:11 I guess ideally they'd get a message notifying them of this when they logged in 03:50:47 is the 34.0 the minor or major version? 03:51:23 major 03:51:35 minor versions are the ones in tag-version.h 03:51:37 I'm thinking that maybe I could update all the versions of a particular verison (0.16) to the latest one right before 0.17 03:52:08 maybe I'd hold off on the latest dev version (0.22) 03:52:22 well, you'd probably just want to update them to the latest point release of 0.17 03:52:25 to take your example 03:52:29 oh, hrm 03:52:31 I guess not 03:52:42 oh, no, yeah that's right 03:52:47 you're talking about 0.16-a 03:53:09 %git 10e7e3d 03:53:09 07gammafunk02 * 0.16-a0-4128-g10e7e3d: Make catoblepas speed 10 and reduce their breath spam a bit (minmay) 10(2 years, 11 months ago, 2 files, 2+ 2-) 13https://github.com/crawl/crawl/commit/10e7e3d4ad27 03:53:12 that's my last 0.16 03:53:40 so you'd update this one to the latest 0.17? i.e. 03:53:45 I don't think that's right... 03:53:46 %git d80ca38 03:53:46 07MarvinPA02 * 0.17-a0-2231-gd80ca38: Remove some cases of vault-specified monster spellsets 10(2 years, 3 months ago, 3 files, 5+ 58-) 13https://github.com/crawl/crawl/commit/d80ca38ea2b9 03:53:57 %git 0.17.0~1 03:53:57 07gammafunk02 * 0.17-b1-37-g455c5dd: Some final 0.17 changelog additions 10(2 years, 2 months ago, 1 file, 3+ 0-) 13https://github.com/crawl/crawl/commit/455c5dd0277a 03:54:02 %git 0.17-b1 03:54:02 07gammafunk02 * 0.17-b1: List the final 0.17 version in the changelog 10(2 years, 3 months ago, 1 file, 2+ 2-) 13https://github.com/crawl/crawl/commit/b0deb92d8f73 03:54:02 oh hm 03:54:06 %git 0.17-b1~1 03:54:07 07gammafunk02 * 0.17-a0-2234-g0da9ddf: In the release guide, use -a in the example git tag command 10(2 years, 3 months ago, 1 file, 1+ 1-) 13https://github.com/crawl/crawl/commit/0da9ddf72ea6 03:54:15 %git 0.16-b1~1 03:54:15 07MarvinPA02 * 0.16-a0-4134-g2ca243f: Let Ashenzari's skill boost handle unrandart staves 10(2 years, 11 months ago, 1 file, 21+ 8-) 13https://github.com/crawl/crawl/commit/2ca243f6681d 03:54:22 could get messy if I already cleaned out the version I wanted to upgrade to 03:54:23 shoudl be that 03:54:29 but I suppose I could just rebuild it 03:54:31 or really just 03:54:42 %git 0.17.0~1 03:54:42 07gammafunk02 * 0.17-b1-37-g455c5dd: Some final 0.17 changelog additions 10(2 years, 2 months ago, 1 file, 3+ 0-) 13https://github.com/crawl/crawl/commit/455c5dd0277a 03:54:50 sorry 03:54:53 %git 0.16.0~1 03:54:53 07gammafunk02 * 0.16-b1-54-gf517508: Update the debian directory for 0.16.0 10(2 years, 10 months ago, 1 file, 6+ 0-) 13https://github.com/crawl/crawl/commit/f517508de3aa 03:54:57 that commit 03:55:02 they would still be alpha 03:55:11 and have everything in 0.16.0 03:55:12 so all 0.16 could go to that one, b1-54 03:55:30 yeah |amethyst would be a good person to talk to about this 03:56:00 yea definitely 03:56:12 nothing stops you from just transfering everyone all the way to before the tag major increase 03:56:13 not going to worry now, but I bet I could free up a LOT of space 03:56:20 but arguably the game would be weird for them 03:56:37 but what you're proposing is probably a nicer thing to do 03:56:45 where is the major tag increase? 03:56:50 all of the versions look like 0.34 03:56:56 how do you mean where? 03:56:57 er. 34.0 03:57:02 we've been on 34 for a long time 03:57:05 yeah 03:57:08 that's what I thought 03:57:25 that's a field, right? 03:57:25 so what do you mean by transferring everyone all the way up to before the tag major increase? 03:57:25 cv, i suppose 03:57:26 no 03:57:27 (i.e. what is the tag major increase) 03:57:38 (vocab lesson) 03:57:43 !lg * x=vsav 03:57:44 8339763. [vsav=34.197] RollingDesktops the Slasher (L6 GnCK of Lugonu), slain by an ogre (a +0 giant spiked club) on D:4 on 2018-01-17 08:57:03, with 324 points after 3517 turns and 0:12:31. 03:57:55 ?/vsav 03:57:55 No matches. 03:58:05 !lg * vsav~~^33 03:58:06 No games for * (vsav~~^33). 03:58:33 !lg * s=regexp_replace(vsav, '\..*', '') 03:58:37 8339763 games for *: 6479929x, 1859834x 34 03:58:48 guess we've not had it as a field before 34 03:59:38 maybe I should have asked, when was our last tag major increase 03:59:58 (thought that a major tag increase would be 35, which hasn't happened yet) 04:00:11 (thought that the *next* increase) 04:00:25 advil wrote a thing about the history of major versions 04:00:29 but i don't remember how to find it 04:01:09 %git 52b7544e8eed 04:01:10 07kilobyte02 * 0.8.0-a0-4443-g52b7544: Bump the major version, purge shims, fix breakage. 10(7 years ago, 17 files, 22+ 869-) 13https://github.com/crawl/crawl/commit/52b7544e8eed 04:01:13 ah, it's in the irc logs: https://gist.github.com/rawlins/6ff153b987756ad8c754a7ee4f735642 04:01:23 that might be the last one 04:01:26 but that's how it's changed 04:01:56 looks like that was to 32 04:03:39 since all my versions are 34, were you suggesting that I could upgrade them ALL to the same single version? 04:04:14 which I could continue to do up until the version just prior to the next tag update, 35, at some indeterminant date in the future? 04:04:29 johnstein: i think gammafunk was imagining you had some versions prior to 34 04:04:38 got it 04:04:42 you certainly *could* require everyone to always update trunk; i understand cpo does that, maybe? 04:04:42 no, I think you're confused as to what that number means/does 04:04:54 you can't transfer games to e.g. 35 04:05:05 yup 04:05:12 once a game is started, it's major version number can never be modified 04:05:24 when we move to 35, we can remove all the minor tags 04:05:33 yea 04:06:07 yeah, actually, I see what you're asking, and yes 04:06:10 heh 04:06:30 if some where 33 or lower then they couldn't upgrade 04:06:36 I'll bug |amethyst. I agree with your comment that upgrading each to the last possible version for each version is nicer and makes more sense 04:06:46 yea, that makes sense. I lucked out 04:07:05 and since I only have to do this once, I'm ok if the process is fiddly and non-repeatable 04:07:12 er. not-automated, I mean 04:08:00 I like this entry in that log advil made 04:08:22 29 Jul 2010 0.8a map layers af76f809 04:08:31 map layers...we must have implemented the z-axis 04:09:18 %git af76f809 04:09:18 07Luca Barbieri02 {by} * 0.8.0-a0-522-gaf76f80: Make map_cell store feat+cloud+item+mons instead of only the topmost one (BREAKS SAVES) (v3) 10(8 years ago, 25 files, 629+ 368-) 13https://github.com/crawl/crawl/commit/af76f809d3f1 04:09:29 looks like there's a savegame-transfer-all.sh function for dgl 04:09:39 oh, and savegame-transfer.sh 04:11:57 nice, a script to transfer..oh right there has to be 04:12:07 that's the script run when the player chooses to transfer their game 04:12:25 ah ok 04:12:27 doh 04:12:35 I'll probably have to cobble something together 04:13:00 I'll start in the ancient versions 04:13:23 %git 5545dd3 04:13:24 07Grunt02 * 0.14-a0-3499-g5545dd3: Fix arc blade discharge behaviour. 10(3 years, 10 months ago, 1 file, 12+ 1-) 13https://github.com/crawl/crawl/commit/5545dd3c9da8 04:13:31 well, that transfer all script might be what you want, yeah 04:13:51 and you can certainly adapt one of those I'm sure 04:14:02 maybe it's even general enough to allow specifying the revision 04:14:38 %git 0.14.1 04:14:38 07Grunt02 * 0.14.1: Changelong for 0.14.1. 10(3 years, 9 months ago, 1 file, 12+ 0-) 13https://github.com/crawl/crawl/commit/7225c62f1965 04:14:53 maybe that's the one I want? i.e. the last tagged version? 04:14:56 -!- amalloy is now known as amalloy_ 04:15:00 (you probably said that above) 04:16:57 hmmm 04:21:46 no, I think you'd want alpha games to stay alpha 04:22:05 if you upgrade them to one of those versions they started alpha but end non-alpha 04:22:35 which matters more for milestones than it does for logfile I guess 04:23:31 -!- amalloy_ is now known as amalloy 04:34:48 03advil02 {gammafunk} 07[stone_soup-0.21] * 0.21.0-4-g039f523: Improve support for MSYS2 10(5 days ago, 1 file, 3+ 3-) 13https://github.com/crawl/crawl/commit/039f52311310 04:34:48 03advil02 {gammafunk} 07[stone_soup-0.21] * 0.21.0-5-g60874cf: Update submodule 10(5 days ago, 1 file, 1+ 1-) 13https://github.com/crawl/crawl/commit/60874cfb0a61 04:34:48 03Aidan Holm02 {gammafunk} 07[stone_soup-0.21] * 0.21.0-6-g5b6045b: Add some new title artwork (froggy) 10(4 days ago, 5 files, 0+ 0-) 13https://github.com/crawl/crawl/commit/5b6045bb1410 04:34:48 03advil02 {gammafunk} 07[stone_soup-0.21] * 0.21.0-7-g05eff09: Tweak build instructions for msys2 10(7 days ago, 1 file, 11+ 1-) 13https://github.com/crawl/crawl/commit/05eff09ddb5b 04:34:48 03MarvinPA02 {gammafunk} 07[stone_soup-0.21] * 0.21.0-8-gf18e19d: Allow manually cancelling Formicid digging 10(8 days ago, 2 files, 6+ 2-) 13https://github.com/crawl/crawl/commit/f18e19db4aea 04:42:14 aidanh: you forgot to cherry-pick that artwork to stable 04:42:22 I forgive you....this time.... 04:43:36 also there are other commits we could probably cherry-pick but I didn't do any that looked like a potential problem I couldn't handle easilly 05:09:44 Stable (0.21) branch on crawl.jorgrun.rocks updated to: 0.21.0-8-gf18e19db4a 05:27:47 -!- amalloy is now known as amalloy_ 05:55:28 Oh drat 05:55:51 Good thing you caught that 05:58:09 I'll go through my commits a bit later tonight and cherry pick some of them; been meaning to do so for a while 06:08:07 The build was canceled. (master - 605b837 #9195 : advil): https://travis-ci.org/crawl/crawl/builds/329707195 09:28:57 -!- rax_ is now known as rax 09:54:30 does anyone have a completed and working apache entry that I can look at? I can't tell what I'm supposed to substitute on the mentioned lines 09:54:30 Huene`: You have 1 message. Use !messages to read it. 09:54:36 !messages 09:54:37 (1/1) espais said (6h 59m 45s ago): any luck with the dgl install after i took off? 09:54:53 !tell espais it's running now but still a work in progress 09:54:54 Huene`: OK, I'll let espais know. 10:04:54 "a stickied post explaining the issue with the bug" 10:04:56 way to be specific 10:05:10 I guess they most likely mean the numlock bug though 10:06:13 hehe yeah that is a bit vague 10:06:57 amalloy_: did you get a chance to test numlock-fix in linux tiles? 10:07:50 nampad-fix, sorry 10:08:26 so if I want to make a manual change to the code on a dgl setup how do I get it to be preserved when an update is done? 10:11:25 as it stands right now when my update runs every night it fails to work afterwards and I have to go make the change again 10:13:10 it seems like you want a script to apply a custom patch, that you know will apply cleanly? 10:13:54 nothing like that will work forever without manual intervention 10:14:20 this will I just need to add imports at the top 10:14:51 oh, is this tornado issues? 10:15:03 yup ascii isn't loaded 10:16:04 well, in this case it won't work forever because eventually someone will get sick of this dependency situation and get things to work on current tornado :D 10:16:56 I'm looking forward to it, that doesn't help now though :) 10:19:39 I don't know how the dgl scripts work exactly, but I think what gammafunk suggested last night is right: (i) construct the relevant patch, (ii) put it somewhere accessible to the dgl scripts, and (iii) modify whatever script does the git pull to immediately apply the patch. I'd probably use "git apply" to apply the patch, maybe using the dry-run mode (git apply --check) first 10:20:23 I'll do that do you happen to know a good resource for patch files for idiots? :D 10:20:32 I have no idea how they work 10:22:19 have you ever looked at the output of git diff? 10:22:43 that's patch format, basically (though there are more bells and whistles for git patches that you probably don't need) 10:23:09 not raw, I have but usually grep'd 10:23:13 oh so they're generated? 10:23:16 the simplest way to get the patch file is to: working in a clean repo, make just the necessary change, don't commit anything. Then use git diff to generate the patch, e.g. "git diff > tornado_ascii_fix.patch". 10:23:20 I always thought I had to write them manually 10:23:46 heh in the old days maybe, but writing patches manually is a huge pain, because of the chunk codes 10:23:53 oh hell that's super easy then 10:24:51 so summary is copy the file I'm changing, change them, diff them output to patch file, git apply --check file.patch, if everything looks good git apply file.patch? 10:25:10 well, don't copy anything, just work in a clean repo on the file that needs changing 10:25:25 what do I compare it against then? 10:25:31 -!- Vaevictu1 is now known as Vaevictus 10:25:31 or does that happen with git apply? 10:26:09 when generating a patch, if you have unstaged changes, "git diff" will automatically produce a patch against the relevant files in master 10:26:27 when applying a patch, it will apply it against the current checked out state (which will probably be master as well) 10:26:31 using "git apply" 10:26:42 so change the file git diff, copy the patch to a safe place, git apply --check make sure it looks good then git apply? 10:28:09 yes...if you want after generating the patch you can do "git checkout -- whatever.py" to reset it in the repo, and then manually test the git apply commands at the command line to make sure you understand what it's doing 10:29:06 (where whatever.py is the file you changed) 10:30:45 will git diff apply all file changes to one patch? 10:30:48 I have to modify two 10:30:58 well 3 if I count another change I want to make 10:31:11 er 1 patch file 10:32:03 you can set it up with multiple patches or just one patch 10:32:19 I'd probably prefer to keep the patches as simple as possible though 10:32:26 one is preferred for my case so like you said above then? 10:33:43 yeah, the steps I described above will work for multiple changed files 10:52:27 publish does copy the contents of webserver to crawl too doesn't it? 10:53:01 the patch is applying in crawl-dev's directory, but not showing up in the active directory, when I publish 10:53:49 not sure how that part of it works 10:53:55 fair enough 10:56:53 -!- amalloy_ is now known as amalloy 10:57:30 advil: i didn't yet. i got halfway through it (tested that things work in linux before your change), but went home before i finished testing post-change 10:57:45 ah ok 11:08:10 I don't get it, my update script runs overnight calling dgl, and publishing, and it breaks ascii, I call the same script after editing crawl-dev's webserver files and it doesn't fix it 11:09:27 possibly stupid question, have you double checked that you are applying the patch after git pull and before any make commands? 11:11:16 https://pastebin.com/WGFHwA8q 11:11:18 that's what I'm doing 11:11:25