00:54:09 -!- Amnesiac_ is now known as Amnesiac 00:57:19 New branch created: pull/1368 (1 commit) 13https://github.com/crawl/crawl/pull/1368 00:57:19 03Alex Jurkiewicz02 07https://github.com/crawl/crawl/pull/1368 * 0.25-a0-805-g2416e7d: Add util/tag-35-upgrade.py (WIP) 10(7 minutes ago, 1 file, 33+ 0-) 13https://github.com/crawl/crawl/commit/2416e7d7e32a 01:14:21 -!- misha is now known as werekitten 02:10:02 Windows builds of master branch on crawl.develz.org updated to: 0.25-a0-804-g50c693f418 02:45:03 Monster database of master branch on crawl.develz.org updated to: 0.25-a0-651-g8c222c7a2b 03:13:38 Unstable branch on crawl.beRotato.org updated to: 0.25-a0-804-g50c693f (34) 07:34:01 New branch created: pull/1369 (2 commits) 13https://github.com/crawl/crawl/pull/1369 07:34:02 03Peter Gerlagh02 07https://github.com/crawl/crawl/pull/1369 * 0.25-a0-783-g73ae0ba: Allow any floor to be annotated with ! 10(9 minutes ago, 5 files, 9+ 50-) 13https://github.com/crawl/crawl/commit/73ae0bab1118 07:34:02 03Peter Gerlagh02 07https://github.com/crawl/crawl/pull/1369 * 0.25-a0-784-gb89853c: I forgot to include files in the previous commit 10(6 minutes ago, 2 files, 257+ 0-) 13https://github.com/crawl/crawl/commit/b89853c09dde 07:42:12 The pull request is still work in progress 07:42:34 I thought it might be useful to put it out there 07:43:10 in case anyone else wants to provide feedback or help me fix the remaining issues 07:47:13 One of the issues I had is that I don't know how to test what a small bit of code does. 09:24:45 can you expand on what you mean by test? 09:24:55 the usual crawldev testing MO is to use wizmode heavily 09:27:05 not sure if it feels strongly to me like that needs a new .cc/.h, dgn-overview.cc is already largely about annotation 10:00:01 03Alex Jurkiewicz02 {advil} 07* 0.25-a0-805-gd17ee4b: Upload release builds to GitHub releases 10(6 days ago, 6 files, 156+ 37-) 13https://github.com/crawl/crawl/commit/d17ee4b226b8 10:00:01 03Alex Jurkiewicz02 {advil} 07* 0.25-a0-806-g520075d: Update release process 10(2 days ago, 1 file, 17+ 46-) 13https://github.com/crawl/crawl/commit/520075d945c5 10:00:01 03Alex Jurkiewicz02 {advil} 07* 0.25-a0-807-g449611f: Simplify CI version detection 10(33 hours ago, 1 file, 12+ 48-) 13https://github.com/crawl/crawl/commit/449611fb78e7 10:00:01 03Alex Jurkiewicz02 {advil} 07* 0.25-a0-808-g03277fe: Remove unnecessary continuation 10(29 hours ago, 1 file, 1+ 1-) 13https://github.com/crawl/crawl/commit/03277fe0e5d3 10:00:01 03Goratrix02 {advil} 07* 0.25-a0-809-g363036d: WIndows installer package build cleanup 10(29 hours ago, 2 files, 2+ 1-) 13https://github.com/crawl/crawl/commit/363036d65451 10:00:01 03advil02 07* 0.25-a0-810-gd808e83: A few more windows build cleanups (Goratrix) 10(19 minutes ago, 2 files, 9+ 3-) 13https://github.com/crawl/crawl/commit/d808e8321f57 10:00:01 03advil02 07* 0.25-a0-811-g2d4e2cf: Simplify mac zip naming 10(3 minutes ago, 1 file, 0+ 12-) 13https://github.com/crawl/crawl/commit/2d4e2cfd7f76 10:00:20 i'd rather have it in a separate set of files (and suggested this initially) 10:00:28 aidanh why? 10:00:46 the dungeon overview has a lot of stuff in it, at least conceptually 10:00:52 in any case if the file is named "annotate.cc" then 80% of dgn-overview.cc needs to be moved into it 10:01:21 I mean, I would support renaming dgn-overview.cc to annotate.cc but 10:01:37 although tbh I wasn't aware of how much of dngn-overview was annotation specific 10:02:03 I guess maybe the intuition was about player-driven annotations? 10:02:15 but most of that file is to support a popup called "Dungeon Overview and Level Annotations" 10:02:31 Unstable branch on crawl.kelbi.org updated to: 0.25-a0-811-g2d4e2cfd7f (34) 10:03:01 it also doesn't have that much in it in the scale of things, barely over 1k lines 10:04:39 ah well that's probably alright then; I'm out at the moment but will take a look when I'm back 10:05:47 yeah I'm guessing it's partly that we call things annotations both that are annotated by the game automatically, and things like this where the player has done something manual, so what you had in mind was a file for the player stuff 10:05:47 which could make sense 10:06:57 For example, when I'm not sure what type something is, I don't know an easy way to find out. I can guess and attempt to use it in the code, but I'll only find out if I was wrong when I compile. For small code changes I can compile, then run Crawl, then test it, but that's quite a lot of work. Compared with what I'm used to in an interactive python 10:06:58 computing environment. 10:08:57 ah, I see 10:09:09 there are IDEs for c++ that will keep track of that for you, so that is an option 10:09:46 but part of the challenge may simply be that when switching from duck typing to something like c++ there is a bunch of stuff you have to keep track of 10:10:00 to some extent in modern c++ you can use `auto` to avoid this in small cases 10:10:17 it's not always good practice though 10:11:28 I actually meant both player annotations and automatic annotations; they're pretty similar apart from how they're created, no? 10:11:34 right 10:11:48 that's why they're all together in dgn-overview.cc :D 10:12:01 sorry, I'm just giving you a hard time at this point 10:12:17 that file does have the ui stuff in it too 10:12:23 so maybe that's what you had in mind 10:13:21 dngn-overview is only annotations from line 840 on actuallly 10:13:26 heh, all good :) 10:14:15 It would be easier for me to move everything into dgn-overview.cc, rather than to figure out how to improve my change to dgn-overview.h (which feels like a botch). But I could also see the merit of separating out the annotations. 10:14:32 I actually don't see anything in it that I wouldn't have described as an annotation or annotation-related 10:14:56 maybe enter_branch, whatever that does 10:15:23 dgn-overview.cc deals with exclusions, and the way exclusions are represented in the annotations 10:15:38 petercordia for working in c++ vs. python, especially in a big project, there is definitely a greater mental overhead in keeping track of things like types, which are also generally much more complicated (though the compiler will catch many issues that you only find at runtime in weird cases in python) 10:16:38 hm, a bunch of it seems travel-cache related (ew) 10:16:38 also tbh I'm not a huge fan of the way crawlcode sometimes uses typedefs for relatively simple data structures, including exactly the case of `typedef map annotation_map_type` that seems to have given you trouble 10:16:38 I think the travel cache is unfortunately intertwined with certain kinds of annotations 10:16:48 (or what I would have called annotations) 10:17:08 because you can autotravel to them 10:18:14 i think it's worth separating the travel-target component of annotations from the annotation part 10:18:17 basically I think that just using `map` as the type is a lot more transparent than forcing the reader to continually keep cross-checking what annotation_map_type is supposed to do. To me this practice feels a bit more C-ish 10:18:18 if that makes sense 10:18:31 yeah, that probably does make sense 10:18:41 I would guess that some aspects of travel targets are in some other file too 10:19:15 (this practice = continually typedef'ing instantiated library template types) 10:19:32 so if you find that confusing, I'm with you 10:23:50 ideally other bits of crawlcode wouldn't have to wrangle the map directly anyway 10:24:43 so typedef'ing it is really just papering over the deeper issue 10:29:19 that sounds like it supports keeping all (automatic and manual) annotations in the same file? 10:33:20 petercordia: sorry, not sure what you mean by that 10:33:51 dungeon annotations (manual and automatic) are kept in the same data structure, yes 10:37:35 I thought you were saying that having different files wrangle the same map-type data structure (annotation_map_type in this case) is the deeper issue, and should be avoided 10:38:21 that's probably true 10:38:25 it's also non-ideal that it is a global static variable 10:38:42 crawlcode is very "cross-loaded" components interact with eachother and there are a lot of globals 10:45:49 right, multiple files accessing the same static data structure (of any type) is a case of multiple components accessing each other's data directly, since components are usually one-per-file 11:09:06 advil: I'm testing the new Makefile in MSYS2 locally now, everything looks to be working fine. The only small issue is that the 'ln -sf' command creates a copy of the file. 11:09:20 I honestly have no idea what the interaction between the unix symbolic links and the windows file system is supposed to be, so I guess this is default behavior. 11:09:51 so it ends up with two idential files named stone_soup-0.25-a0-811-g2d4e2cfd7f-console-win32.zip and stone_soup-latest-console-win32.zip 11:14:18 windows doesn't support symlinks right? 11:15:18 not sure, googled it, and it may support, but I never used them 11:15:58 I also found a screenshot showing there is checkbox that can be selected when *installing* MSYS2 to enable support of symlinks 11:16:25 which one is the new makefile btw 11:16:25 looks like some hacks are needed 11:17:02 the one in master, there were changes for the automated github packaging 11:17:26 oh nice, I've been away for too long I see 11:23:57 I don't think windows supports symbolic links at all? 11:24:17 the use of ln there did get a bit of side-eye from me but I ignored it :D 11:24:35 oh I didn't read the messages since my ping 11:24:45 looks like it does, but I have never seen them used: https://blogs.windows.com/windowsdeveloper/2016/12/02/symlinks-windows-10/ 11:26:35 yesh, the whole thing seems to be really hacky: https://www.joshkel.com/2018/01/18/symlinks-in-windows/ 11:27:18 so probably it's best to ignore it 11:27:18 yeah 11:29:03 I think the use of ln there may really be aimed specifically at the case of cross-compiling during the gh action, so maybe it doesn't need to always happen 11:29:35 basically so the files to upload have convenient stable names like `crawl-ref/source/mac-app-zips/latest.zip` 11:30:05 yes, makes sense, if it could be conditionally excluded when doing a local build 11:30:25 or just on windows 11:32:44 yep 11:34:41 it would also be best to have some clean target that removes this stuff (which I guess can't be `clean` the way this is set up) 11:54:09 03Umer Shaikh02 {ebering} 07* 0.25-a0-812-g0ddb669: Refactor some regen code 10(9 weeks ago, 3 files, 30+ 59-) 13https://github.com/crawl/crawl/commit/0ddb669d6f98 11:54:09 03Umer Shaikh02 {ebering} 07* 0.25-a0-813-g4828e68: Make TLA and regen artefacts activate after maxing hp 10(9 weeks ago, 4 files, 147+ 46-) 13https://github.com/crawl/crawl/commit/4828e68a6e45 11:54:09 03Umer Shaikh02 {ebering} 07* 0.25-a0-814-g5fea3be: Save compatibility for changes to regen items 10(7 weeks ago, 2 files, 14+ 0-) 13https://github.com/crawl/crawl/commit/5fea3be28908 12:27:06 Unstable branch on crawl.kelbi.org updated to: 0.25-a0-814-g5fea3be289 (34) 12:31:21 Anyone has an idea why the gitignore entry '/source/stone_soup' doesn't ignore the zipped packages for me? On Windows 10, using VSCode and Git. I could swear it worked fine yesterday. Changing it to ignore 'source/stone-soup*.zip' works again 12:31:54 one is stone_soup, the other is stone-soup ? 12:32:32 sorry, typo 12:32:32 second one is _ too 12:32:37 file name is stone_soup-latest-tiles-win32.zip for example 12:32:57 and having /source/stone_soup in the gitignore doesn't ignore it 12:34:34 i think that's expected? i.e. if there was a directory or a file named stone_soup, it'd be ignored, but not a file merely prefixed with stone_soup? 12:34:50 oh ok 12:35:07 I thought it's supposed to check file names too 12:36:21 <|amethyst> if you don't use a wildcard, it's an exact match 12:37:09 ok, I'll do a PR shortly to change it to source/stone_soup*.zip 12:37:13 thanks :) 12:39:00 oh, looks like that travel cache stuff with anntations is just for transporters 12:39:06 !blame3 gammafunk 12:39:06 gaaaaammaaaaafuuuuunk 12:43:23 hm, and runed doors too apparently 12:44:53 aidanh: what's the problem with annotations? 12:46:14 and yeah, the extension of travel cache annotations to transporters was based off the runed door implementation 12:46:28 gammafunk: nothing especially; we were just discussing the contents of dgn-overview.cc and the topic of the travel-cache came up 12:46:31 if you want autoexplore warnings to tell you about untaken transporters, you need something like that, as I recall 12:46:54 ahh fair enough, makes sense 12:47:01 hrm, or maybe it's just so you can get the annotations on ctrl-o; it's been so long since I implemented that! 12:47:16 but I do recall autoexplore actually checks for presence of annotations in some way 12:48:38 New branch created: pull/1370 (1 commit) 13https://github.com/crawl/crawl/pull/1370 12:48:38 03Goratrix02 07https://github.com/crawl/crawl/pull/1370 * 0.25-a0-812-g8120fc8: Fix gitignore for windows packages 10(6 minutes ago, 1 file, 2+ 1-) 13https://github.com/crawl/crawl/commit/8120fc8466e2 12:51:20 03Alex Jurkiewicz02 {Aidan Holm} 07* 0.25-a0-815-gab441bd: Always save macros 10(7 days ago, 2 files, 2+ 7-) 13https://github.com/crawl/crawl/commit/ab441bd002d5 12:52:25 Thrustazar (L18 DrAE) Crash caused by signal #15: Terminated (Swamp:3) 13:02:42 Unstable branch on crawl.kelbi.org updated to: 0.25-a0-815-gab441bd002 (34) 13:04:24 -!- raskol_ is now known as raskol 13:12:13 halsver (L22 DgAs) ASSERT(mgrd(mons->pos()) == s) in 'show.cc' at line 275 failed. (Abyss:1) 13:32:05 Unstable branch on crawl.akrasiac.org updated to: 0.25-a0-815-gab441bd (34) 14:32:38 !crashlog halsver 14:32:39 1. halsver, XL22 DgAs, T:99688 (milestone): http://crawl.berotato.org/crawl/morgue/halsver/crash-halsver-20200419-171212.txt 14:32:43 !crashlog Thrustazar 14:32:44 2. Thrustazar, XL18 DrAE, T:37011 (milestone): http://crawl.akrasiac.org/rawdata/Thrustazar/crash-Thrustazar-20200419-165225.txt 14:34:56 I think the first one may have been fixed (though the bug that was directly fixed involved a more, which I don't see there) 14:34:56 %git 7dc260b4ef 14:34:56 07Aidan Holm02 * 0.25-a0-436-g7dc260b: Fix more() being called while swapping monsters 10(3 months ago, 2 files, 24+ 6-) 13https://github.com/crawl/crawl/commit/7dc260b4ef9e 15:26:17 -!- aidanh_ is now known as aidanh 18:22:19 Unstable branch on underhound.eu updated to: 0.25-a0-815-gab441bd002 (34) 18:43:43 03nirrattner02 {ebering} 07* 0.25-a0-816-g270f1c6: Prevent overtraining skill targets 10(5 months ago, 4 files, 63+ 7-) 13https://github.com/crawl/crawl/commit/270f1c643a1f 18:47:02 03Alex Jurkiewicz02 {ebering} 07* 0.25-a0-817-geca56f1: Make species-gen.py work with Python 2 & 3 10(3 months ago, 3 files, 26+ 11-) 13https://github.com/crawl/crawl/commit/eca56f19eff3 18:47:02 03Alex Jurkiewicz02 {ebering} 07* 0.25-a0-818-g2ff8117: ci: Install PyYAML for Python 3 on macOS 10(8 weeks ago, 1 file, 1+ 2-) 13https://github.com/crawl/crawl/commit/2ff811776227 18:50:11 Unstable branch on crawl.kelbi.org updated to: 0.25-a0-816-g270f1c643a (34) 19:02:51 Unstable branch on crawl.kelbi.org updated to: 0.25-a0-818-g2ff8117762 (34) 19:05:06 advil: if you want to test that ci code, you could backport it to the 0.24 branch and then create a new 0.24 release 19:05:25 (I think there were some commits on that branch not yet tagged anyway) 19:07:21 about the ln for package names. I think the easiest solution may be to write the zip to a static file name. The local build doesn't need to have a version in the filename ever 19:07:43 especially now that the builds getting shared are made automatically. developers aren't making builds to share any more (theoretically!) 19:08:07 eg always write to dcss-win32-tiles.zip with no version 21:02:47 UpsideDownChuck (L24 VpIE) ERROR in 'skills.cc' at line 1949: ASSERT failed: lev of 28 out of range 0 (0) .. MAX_SKILL_LEVEL + 1 (28) (Vaults:3) 21:14:11 good error 21:15:04 !crashlog 21:15:28 21890. nxtlvl, XL27 KoFi, T:50823 (milestone): https://crawl.kelbi.org/crawl/morgue/nxtlvl/crash-nxtlvl-20200419-181221.txt 21:19:38 !crashlog UpsideDownChuck 21:19:38 No milestones for UpsideDownChuck (crash). 21:20:39 excuse you sequell 21:20:47 !crashlog UpsideDownChuck 21:20:47 No milestones for UpsideDownChuck (crash). 21:23:35 !crashlog UpsideDownChuck 21:23:36 No milestones for UpsideDownChuck (crash). 21:23:39 argh 21:24:04 https://crawl.kelbi.org/crawl/morgue/UpsideDownChuck/crash-UpsideDownChuck-20200420-010232.txt 21:55:57 03ebering02 07* 0.25-a0-819-gc0ddf3c: Don't crash when skill target set to 27 10(80 seconds ago, 1 file, 13+ 7-) 13https://github.com/crawl/crawl/commit/c0ddf3c23efd 22:03:35 Unstable branch on crawl.kelbi.org updated to: 0.25-a0-819-gc0ddf3c23e (34) 22:28:55 I have a bunch of corrections to TAG_MAJOR_VERSION logic. Should I bundle them into one commit or split into a few ones separated by issue? 22:28:55 17 files changed, 311 insertions(+), 29 deletions(-) 22:30:06 -!- Tux[Qyou] is now known as werecarpet 23:30:12 probably split them out, but I also don't know what you mean by tag major version logic 23:30:22 instances of use of tag major that are incorrect? 23:31:05 those do have to be looked at carefully; sometimes sort of strange code blocks have to be set up chaining multiple uses 23:31:15 and it can seem weird or incorrect at first glance