00:51:33 <10P​leasingFungus> hm, i don't see a pull request 00:52:07 <10P​leasingFungus> oh, it was closed 00:52:23 <10P​leasingFungus> after just a few hours...? feels quite sudden 01:08:46 03Umer Shaikh02 {PleasingFungus} 07* 0.26-a0-671-gab20509: Check right god for life-saving in god description 10(8 days ago, 3 files, 70+ 39-) 13https://github.com/crawl/crawl/commit/ab205098cb12 01:08:46 03PleasingFungus02 07* 0.26-a0-672-g1df16c3: Rephrase lifesaving desc 10(57 seconds ago, 1 file, 1+ 1-) 13https://github.com/crawl/crawl/commit/1df16c338629 01:22:51 Unstable branch on crawl.kelbi.org updated to: 0.26-a0-672-g1df16c3386 (34) 01:23:44 -!- amalloy is now known as amalloy_ 01:34:29 Unstable branch on crawl.develz.org updated to: 0.26-a0-672-g1df16c3386 (34) 01:55:19 Windows builds of master branch on crawl.develz.org updated to: 0.26-a0-672-g1df16c3386 02:26:56 Unstable branch on cbro.berotato.org updated to: 0.26-a0-672-g1df16c3386 (34) 02:53:36 Monster database of master branch on crawl.develz.org updated to: 0.26-a0-672-g1df16c3386 03:30:58 Fork (bcrawl) on crawl.kelbi.org updated to: 0.23-a0-3603-ge1cc4b0452 03:43:40 Fork (bcadrencrawl) on crawl.kelbi.org updated to: 0.22.1-3068-g3bd91d997f 08:51:34 I can't really figure out what problem in testing a mock player class is trying to solve, but it doesn't seem likely that it's going to be my preferred solution 08:54:32 I guess they closed all their PRs and left? 09:00:55 probably got a bit frustrated with the lack of feedback, but they seemed to be looking for a big introductory project and just chose something that's not likely to get as much attention 09:01:26 well, they also ignored a bunch of feedback they did get 09:01:40 someone may want to look at #1587 still 09:02:21 the actual tests are mostly not what I would be inclined to do, but there is currently not any way to do anything with lua in catch2 09:05:33 are there actual open source projects where PRs do get immediate feedback? 09:05:33 maybe ones with giant dev communities 09:05:33 people seem to expect that a lot 09:05:52 anyways, I'm drowning in obligations right now, so... 09:08:33 yeah, this is definitely not a project where one can expect immediate PR feedback 09:32:50 03Peter Hurst02 {gammafunk} 07* 0.26-a0-673-g44785f1: Add a tag to set vaults as passable 10(1 year, 6 months ago, 3 files, 19+ 3-) 13https://github.com/crawl/crawl/commit/44785f12d51f 09:32:50 03gammafunk02 07* 0.26-a0-674-g18554e4: Add transporter vaults to Vaults 10(5 hours ago, 2 files, 153+ 87-) 13https://github.com/crawl/crawl/commit/18554e4600ae 09:32:50 03gammafunk02 07* 0.26-a0-675-g30791e7: Some transporter vault adjustments 10(5 hours ago, 2 files, 30+ 28-) 13https://github.com/crawl/crawl/commit/30791e7224dc 09:32:50 03gammafunk02 07* 0.26-a0-676-g62c44f1: Adjust doors in a runed door vault 10(5 hours ago, 1 file, 4+ 4-) 13https://github.com/crawl/crawl/commit/62c44f1a976a 09:33:34 wow, didn't realize how many commits advil made this release so far! 09:33:41 he's in the lead by a good amount 09:33:53 huh, am I? 09:34:02 maybe seven years ago you could expect feedback pretty fast 09:34:12 157 advil 09:34:12 120 ebering 09:34:12 111 PleasingFungus 09:34:12 52 Aidan Holm 09:34:12 49 kate- 09:34:14 crawl development has since slowed down because the game is perfect, now. 09:34:15 32 gammafunk 09:34:17 I thought I've only done some bugfixes in master 09:34:33 I wonder if that is somehow counting branches 09:34:47 hrm 09:34:50 did I mess up my command 09:35:25 I guess I did some refactors 09:35:55 oh right, also food stuff 09:36:07 time is confusing 09:36:13 yeah, seems correct 09:36:33 kate has more commits than I do so far, this will not stand 09:51:53 Unstable branch on crawl.kelbi.org updated to: 0.26-a0-676-g62c44f1a97 (34) 09:56:04 I had a brief look at https://github.com/crawl/crawl/pull/1587, it probably needs a bit of polish but it looks like it could be a good start 09:58:26 i'm not so sure that testing via lua is the right way to go, but getting all that env setup stuff should allow more fine-grained testing 09:58:59 (afaik anyway; my impression of lua tests was that it's mostly stress testing, might be completely wrong though) 09:59:31 there's a lot of stuff that is really integration testing as ebering pointed out the other day 10:00:24 i must have missed that 10:01:30 yeah admittedly there aren't many clear 'units', but integration testing still has value 10:01:35 I don't think that the stuff that is currently done as lua integration tests really makes sense to port to catch2 (e.g. there is a bunch of stuff in this PR that generates a line of lua code via make_stringf and then runs it) 10:01:46 but the main gain I see with this PR would be allowing lua code to set up conditions for things that are better done in catch2 10:02:31 well, maybe that's a bad description of what this code does 10:02:46 allowing lua code to set up conditions? i.e. using lua to set up the test scenario, then using C++ to peer into internals that aren't exposed? 10:02:57 yeah 10:03:11 basically like this iterator case that they were asking about 10:03:26 re the wrapper methods this PR adds, I don't think those are a good idea; I think just testing the c++ method directly is fine 10:03:40 (although testing the lua API also has value of course) 10:03:40 or something like moncast.lua which is mostly written in c++ 10:04:29 re peering into internals, generally that's frowned upon in unit testing, but then there are no clear-cut units here anyway... 10:04:43 I think the use of mock frameworks would be a bad idea though 10:05:15 these heavyweight globals are complex enough that mocking would introduce a lot of pain; it's better just to wipe them clean before each test 10:18:34 03Aidan Holm02 07[remove-env-defines] * 0.26-a0-677-g233fbfb: Remove menv() helper macro 10(30 hours ago, 33 files, 117+ 118-) 13https://github.com/crawl/crawl/commit/233fbfb87fec 10:18:34 03Aidan Holm02 07[remove-env-defines] * 0.26-a0-678-gaaf76a3: Remove mitm() helper macro 10(29 hours ago, 54 files, 416+ 417-) 13https://github.com/crawl/crawl/commit/aaf76a3eef98 10:18:34 03Aidan Holm02 07[remove-env-defines] * 0.26-a0-679-g8c6b855: Remove grd() helper macro 10(29 hours ago, 106 files, 817+ 818-) 13https://github.com/crawl/crawl/commit/8c6b85500dc7 10:18:34 03Aidan Holm02 07[remove-env-defines] * 0.26-a0-680-g2114a8c: Remove mgrd() helper macro 10(29 hours ago, 27 files, 85+ 86-) 13https://github.com/crawl/crawl/commit/2114a8c1a14e 10:18:34 03Aidan Holm02 07[remove-env-defines] * 0.26-a0-681-g6e410d8: Remove igrd() helper macro 10(29 hours ago, 20 files, 64+ 67-) 13https://github.com/crawl/crawl/commit/6e410d8402cb 10:22:12 aidanh: I was hoping to read some motivation in that first commit 10:22:22 ...but there is nothing...but removal... 10:22:23 gammafunk: i added some 10:22:26 oh nice 10:22:38 yeah I read that last batch, didn't realize you'd updated 10:22:50 how do you automate these commits btw? 10:23:03 or make them in a sane way, I guess I should ask 10:24:36 I guess most of it is very simple search and replace 10:24:36 the first step is rg '\b(igrd|mgrd|grd|mitm|menv)\b' -l | xargs -o vim 10:24:42 <10P​leasingFungus> aah 10:24:43 that just opens all the files that need to be changed 10:25:26 then it's :bufdo %s/menv/env.mons/g, etc, for each of the patterns, and :wqa 10:25:53 thx 10:27:14 er, that should be :bufdo %s/\/..., so that it only matches at word boundaries 10:35:47 04Build failed for 08remove-env-defines @ 6e410d84 06https://github.com/crawl/crawl/actions/runs/315650201 10:41:35 03amcnicky02 {Aidan Holm} 07* 0.26-a0-677-g3c51cfd: Give Edmund a weaponless tile, fixing #1453 10(5 weeks ago, 3 files, 14+ 0-) 13https://github.com/crawl/crawl/commit/3c51cfd513fd 10:51:12 for purposes of deciding what ought to be done, I'm not sure how to decide what is and isn't internals in crawlcode, it's all internals 10:51:21 except in a few zones of the code 10:53:20 usually the interface of the unit being tested, but they can still be pretty blurry 10:54:15 crawlcode has blurry vision 3 10:54:19 heh 10:54:58 04Build failed for 08remove-env-defines @ 6e410d84 06https://github.com/crawl/crawl/actions/runs/315650201 10:55:19 <10P​leasingFungus> no wonder dev slowed down! 10:56:21 "ModuleNotFoundError: No module named 'yaml'" looks like deps installations on mac os is busted 10:57:39 Unstable branch on crawl.kelbi.org updated to: 0.26-a0-677-g3c51cfd513 (34) 11:44:52 %cache 11:44:52 CCACHE status: Hit Rate =67.01 % | Files Cached=1916 | Cache Size=2.0 GB | Max Cache Size=5.0 GB 11:45:33 03Aidan Holm02 07* 0.26-a0-678-g4ee1432: Remove menv() helper macro 10(31 hours ago, 33 files, 117+ 118-) 13https://github.com/crawl/crawl/commit/4ee14324bdca 11:45:33 03Aidan Holm02 07* 0.26-a0-679-g4c0a3a5: Remove mitm() helper macro 10(31 hours ago, 54 files, 416+ 417-) 13https://github.com/crawl/crawl/commit/4c0a3a5005b8 11:45:33 03Aidan Holm02 07* 0.26-a0-680-ge1d8dbd: Remove grd() helper macro 10(31 hours ago, 106 files, 817+ 818-) 13https://github.com/crawl/crawl/commit/e1d8dbde3443 11:45:33 03Aidan Holm02 07* 0.26-a0-681-gf84b031: Remove mgrd() helper macro 10(31 hours ago, 27 files, 85+ 86-) 13https://github.com/crawl/crawl/commit/f84b0316f654 11:45:33 03Aidan Holm02 07* 0.26-a0-682-g294304f: Remove igrd() helper macro 10(31 hours ago, 20 files, 64+ 67-) 13https://github.com/crawl/crawl/commit/294304f6e872 11:47:34 speaking of ccache, i'm seeing a lot of cache hit rate = 0.00% in CI logs 11:49:16 even for things like https://github.com/crawl/crawl/runs/1271573232 where you'd expect nearly a 100% hit rate 11:50:30 yeah, not sure our CI ccache is actually working 11:51:00 %git bogged_down 11:51:00 Could not find commit bogged_down (git returned 128) 11:51:38 %git origin/bogged_down 11:51:38 Could not find commit origin/bogged_down (git returned 128) 11:51:38 %git bogged-down 11:51:38 07PleasingFungus02 * 0.26-a0-701-g7874d04: Remove insubstantial wisps from swamp gen 10(35 hours ago, 1 file, 0+ 1-) 13https://github.com/crawl/crawl/commit/7874d0465818 11:56:37 <10P​leasingFungus> dogged bown 11:58:14 <03w​heals> truly, pf is now the bog champ 12:24:17 Unstable branch on crawl.kelbi.org updated to: 0.26-a0-682-g294304f6e8 (34) 12:37:06 New branch created: test-iterators (1 commit) 13https://github.com/crawl/crawl/tree/test-iterators 12:37:06 03Aidan Holm02 07[test-iterators] * 0.26-a0-683-gb2a6397: Add some test cases for coordit.cc 10(73 seconds ago, 2 files, 205+ 0-) 13https://github.com/crawl/crawl/commit/b2a6397c1dcc 13:29:52 Unstable branch on crawl.akrasiac.org updated to: 0.26-a0-682-g294304f (34) 15:05:03 -!- aidanh_ is now known as aidanh 16:35:27 I am sorry. I closed those pull requests because I thought they are all unimportant. 16:35:57 One thing though. 16:37:40 I think I read that I should discuss plans for new feature here 16:37:50 before I try to work on it. 16:38:38 but that seems to get burried among the many things that I talked about here. 16:39:15 Is it a good idea to add something like a plan or proposal tag 16:39:19 in the github issue? 16:39:53 CrawlCycle: well, I think there was some discussion about the scope of your changes at least initially, but I wasn't around for all of it 16:40:25 It is helpful to tag things, but I wouldn't say it has a major impact on the workflow or timing of resolving a PR or anything 16:40:49 it is about if i should spend time on a plan 16:41:02 before actually submiting a pull request 16:41:35 i wasn't sure if the lua fixture is a good idea to you guys. 16:41:57 i just found it makes writing tests easier for me. 16:42:16 and it took me quite a long time to work on it. 16:42:27 that is risky in terms of potential efforts wasted. 16:42:51 maybe i should ask about it a few more times here? 16:44:05 another thing is that this is the first time i ever work with others on code 16:44:05 so I don't know what is the expected time required to review a pull request 16:44:05 a plan is a good way to record and brainstorm the rationale, scope, and approach of your changes. If you're not sure of a lot of things and want to centralize the feedback you get for a larger project, a plan is nice; we use them all the time via the wiki github feature 16:44:16 if i know it generally takes a month to review a pull request 16:44:25 that would be fine too. but i didn't know. 16:44:45 erh. wait I can edit the wiki? 16:44:56 for PRs, they often can be reviewed quite quickly if they're small, but I think your project is just an inherently large one and it didn't seem that the people who looked at your proposed changes were sure about the idea 16:45:11 also those people are pretty busy at the moment, so their time to look at crawl things can be sporadic 16:45:25 for making a plan, you can use the wiki pages of your own fork of the repo 16:45:39 and link to it in a PR 16:46:08 what you can sort of do to avoid frustration is have a larger "back burner" project that you iterate on 16:46:35 oh. so you mean to write about it in the wiki of my own fork, and then post the link here? 16:46:42 and have smaller projects that are well-defined and not too contraversial or high impact or high effort 16:47:13 so you can get some things for these smaller projects merged near term and not have to sit on one large project without seeing any results 16:47:55 i nearly thought about drawing a histogram for the expected time to merge pull requests. 16:47:57 so the smaller projects could be simple bugfixes unrelated to your large project, or just basically parallel things like documentation, whatever you want to work on that's not the big project that needs all this thinking and feedback etc 16:48:27 oh. i get it. so you mean instead of using github issues 16:48:36 use the pull request itself to ask for feed back 16:48:44 and before getting any feed back, don't work on it. 16:49:02 is this what you mean? 16:49:41 well I'm talking more about what *other things* you can work on as you take a break from a larger project or are just waiting to hear back from people etc 16:49:48 you can work on smaller scope things that tend to get feedback more quickly and just keep iterating on a large project (like the testing one you started) 16:50:12 I think github issue could be a way to link your wiki doc, yeah 16:51:07 You can make an issue that's about "Testing lua framework" (just to pick an example name) that gives a brief synopsis of the issues and that links to your wiki page, for example 16:51:25 You could probably also brainstorm in the issue, although I'm not sure that is the best format 16:51:38 you can see some wiki pages in our repo if you want examples of how we use those pages 16:51:43 I remember now. I read the contribution-process.md 16:52:10 and it says I should discuss here first. 16:52:41 But overall I'm saying you don't have to put all your effort into this one larger project that needs all this planning; you can come back to the large project as you get feedback, as people who are interested have time to discuss etc 16:53:15 and work on smaller things like bug fixes or whatever else 16:53:57 obviously you can also work on other open source projects you find; just depends on what you see that you'd like to work on 16:54:18 in the crawl codebase, I mean; maybe there aren't other projects that interest you 16:55:17 re feedback, this channel is certainly a starting point, and you have gotten a bunch of discussion. But you can certainly write up a plan for any project ahead of time and start by showing that to people 16:55:17 If you think something that formal is useful 16:57:11 i think github issue works better than a blank pull request for soliciting feedback in some case. 16:57:32 for example, plan: remove global and add constness 16:57:58 refactor by adding defualt argument in the form f(const _you=you) 16:58:34 that kind of thing doesn't have a well define scope. it can happen on any file of crawl. so a pull request might not work so well. 16:59:53 if i talk about it here, i would have to scan the log of the channel to fish out feed back by various people 17:00:03 that is kind of inefficient. 17:00:53 the drawback is that it might clutter the issue page of github 17:00:53 i hope it is acceptable to you guys. 17:02:25 well issues are like bug reports; it's expected that there are a lot of them for a large project like this, so the number is not too big a concern. But you could have one issue for collecting them, and just catalog them in the issue 17:03:09 one issue for collecting all feedbacks. 17:03:27 you see. if i talk about it here, i can talk about one thing for days. 17:03:39 if it is an github issue, that would translate to 15 edits on a comment 17:03:50 and no one would see pages of text. 17:04:21 unless we have github emails disabled, we get an email for basically every issue/comment 17:04:31 erh. every edit too?? 17:04:40 not for edits 17:04:42 i hope not. that would be annoying. 17:04:54 I'm not sure that is the best way to iterate on something / get feedback though 17:05:29 what is the best way in your mind? 17:05:46 well, posting here probably 17:05:48 fwiw I think much the devteam is largely swamped by non-crawl things right now 17:06:29 i didn't now that. That was why I thought about drawing a histogram. Numbers don't lie. 17:07:13 adding a default player argument to functions seems non-ideal btw 17:07:30 why? 17:08:06 CrawlCycle: If I may, I'm still - even having read most of the discussion here - a bit confused about your overall goal (whereas I could say, for example, that advil and gammafunk are trying to make the game more interesting and challenging, etc) 17:08:42 i want to make the most used code documented and tested. 17:08:55 documented, so that new developer can know what is going on quickly 17:09:09 tested, so that any refactoring would be less risky 17:09:31 refactor to make the code easier to work with without changing the functionality of the code. 17:10:03 adding constness to ensure an argument stays constant if that is the intention. 17:11:04 documentation depends on writing test to figure out the behavior 17:11:14 i am not you guys. i don't know the code well. so i want to test 17:11:18 before writing the doc. 17:12:05 and test means i want to make writing test faster and easier. 17:12:15 I am not one of "you guys", FTAOD; I think I have one patch in vanilla after ebering rewrote it from the inside out. 17:13:59 But I think (for example) that major refactoring has the obvious downside that it will be vexing for the existing developers, who are used to how things are right now; even, perhaps, to people like me with only a moderate familiarity with the source. 17:14:25 i am mostly interested in doc and test. 17:15:01 i wouldn't want to change the code before knowing it well. 17:17:42 advil: why adding that kind of default argument is bad? 17:17:47 i thought it wouldn't break existing api and the change is local 17:18:04 if the constness is wrong, compilation of the code would immediately tell. 17:19:47 after that, people can slowly replace every call of the function 17:19:54 by using the default argument. 17:20:17 so it is an incremental refactor that slowly add constness and limit the use of global 17:20:29 even if someone decide to not do the job at anytime 17:20:48 it won't hurt too much. partial const correctness is better than no const. 17:22:16 in the end. make the default argument an required argument to prevent future violation of constness. 17:28:55 <10P​leasingFungus> seems awkward to have to pass a ‘you’ into many places 17:28:55 <10P​leasingFungus> if i’m understanding the suggestion correctly 17:29:15 <10P​leasingFungus> very unclear that the code would be more readable 17:30:18 that would be a drawnback 17:31:49 in the ideal case, we would organize those function into a class 17:32:10 with the constructor require either you& or const you&, I guess? 17:32:55 <10P​leasingFungus> can you give a snippet of example code? 17:33:05 <10P​leasingFungus> i’m not sure what you’re suggesting 17:33:28 let me open the ide 17:34:48 i have been looking at the actor_near_iterator class for days 17:34:56 so i will use that as an example 17:35:29 actor_near_iterator::actor_near_iterator(coord_def c, los_type los, player& you_, crawl_environment& env_) 17:35:34 was originally 17:35:47 actor_near_iterator(coord_def c, los_type los) 17:36:15 sorry 17:36:54 actor_near_iterator(coord_def c, los_type los, player& you_=you, crawl_environment& env_=env) : _you(you_), _env(env_) 17:37:04 that is the modified version 17:39:00 the correct way is to have both an iterator and a constant iterator. this change only limit the use of global you and global env within the class/function. 17:41:35 i would think that implementing the constant iterator would need a const& env argument though. 17:42:39 anyway, i like doc and test. this is just something for anytime i get bored and want to see quick result. 17:43:14 dead simple refactor. 17:54:41 I'm really having trouble understanding what this would accomplish, but it doesn't seem simpler 17:55:05 except make the function signature harder to understand 17:55:14 this is what i think: 17:55:44 incrementally enforce constness without breaking the api 17:56:05 if the function doesn't change "you", then 17:56:21 the default arguement would be `const you_&=you` 17:56:57 that way, i know for sure that the function won't change you 17:56:57 if the function would change you, then the compilation would fail. 17:57:14 2. help to eventually stop using global 17:57:29 if you want the api for actor_near_iterator to have a const version, then why not just have operator* return a const actor *? 17:57:49 I mean, actor_near_iterator in general needs to have a non-const version 17:58:15 not using globals is a noble ideal but I'm not sure this is the way to move towards it 17:59:30 hmm. usually in examples online, if the whole class doesn't want to change a thing 17:59:46 it takes in the thing by a const reference or by value 17:59:58 that way regardless of the implementation inside 18:00:06 the argument already enforce constness 18:00:52 enforcing constness in every method require every method enforce const correctly 18:00:58 ok, I think we are talking about different things 18:01:27 you want the compiiler to enforce that actor_near_iterator does not change you or env 18:01:28 sorry? i don't get. so what do you mean? 18:01:45 const method, const argument, or const return value? 18:02:33 i was talking const everything starting from const arguments. 18:03:03 advil, yes that is what i want to say. 18:03:44 i don't want to think about implementation detail when i don't know the code well. 18:03:53 add const to everything, if it compiles, great 18:04:02 if it doesn't, then i know i did something wrong. 18:07:14 Before i can ask the compiler to enforce that some class/function to not change env and player 18:07:31 i need to ask the class/function stop using the global env/player 18:08:10 and instead use either a reference or constant reference of env/player. 18:25:16 Unstable branch on underhound.eu updated to: 0.26-a0-682-g294304f6e8 (34) 18:28:51 hmm. i need to work on other things. 18:28:55 see you guys later. 19:29:14 -!- bairyn is now known as ByronJohnson 20:28:03 New branch created: pull/1588 (3 commits) 13https://github.com/crawl/crawl/pull/1588 20:28:03 03CrawlCycle02 07https://github.com/crawl/crawl/pull/1588 * 0.26-a0-666-ga8bc06c: docs: Document invalidate_agrid 10(6 days ago, 1 file, 31+ 5-) 13https://github.com/crawl/crawl/commit/a8bc06c93d5e 20:28:03 03CrawlCycle02 07https://github.com/crawl/crawl/pull/1588 * 0.26-a0-684-ge44aacc: Merge branch 'master' into crawlcycle/doc/2 10(82 minutes ago, 0 files, 0+ 0-) 13https://github.com/crawl/crawl/commit/e44aacceb9bb 20:28:03 03CrawlCycle02 07https://github.com/crawl/crawl/pull/1588 * 0.26-a0-685-g3a393b6: docs: Revise docs as suggested by Pleasing Fungus 10(10 minutes ago, 1 file, 11+ 18-) 13https://github.com/crawl/crawl/commit/3a393b60781f 20:44:52 New branch created: pull/1589 (1 commit) 13https://github.com/crawl/crawl/pull/1589 20:44:52 03CrawlCycle02 07https://github.com/crawl/crawl/pull/1589 * 0.26-a0-683-g614c2df: docs: Document artp_value_type 10(11 minutes ago, 1 file, 6+ 4-) 13https://github.com/crawl/crawl/commit/614c2dffce3c 21:18:30 New branch created: pull/1590 (4 commits) 13https://github.com/crawl/crawl/pull/1590 21:18:30 03CrawlCycle02 07https://github.com/crawl/crawl/pull/1590 * 0.26-a0-670-gd7104af: test: Add catch2 fixture for running dLua/cLua 10(34 hours ago, 6 files, 685+ 1-) 13https://github.com/crawl/crawl/commit/d7104afbdd8b 21:18:30 03CrawlCycle02 07https://github.com/crawl/crawl/pull/1590 * 0.26-a0-671-g591b3c0: style: Remove braces 10(34 hours ago, 1 file, 3+ 4-) 13https://github.com/crawl/crawl/commit/591b3c0109d1 21:18:30 03CrawlCycle02 07https://github.com/crawl/crawl/pull/1590 * 0.26-a0-672-gc690abf: style: Add newline to end of file 10(34 hours ago, 5 files, 5+ 5-) 13https://github.com/crawl/crawl/commit/c690abfafb10 21:18:30 03CrawlCycle02 07https://github.com/crawl/crawl/pull/1590 * 0.26-a0-686-g4e0240f: Merge branch 'master' into crawlcycle/test/2_lua_fixture 10(30 minutes ago, 0 files, 0+ 0-) 13https://github.com/crawl/crawl/commit/4e0240f62667 21:18:44 advil: re what providing dependencies like you/env in constructors would accomplish: this would be useful for tests and would help for a 'libcrawl' 21:25:22 but unit tests can reset the global you instance, so testing is not so much of a problem in practice, and libcrawl is not really a priority 21:46:31 03CrawlCycle02 07https://github.com/crawl/crawl/pull/1590 * 0.26-a0-687-gc08ff27: fix: Stop using the igrd and mgrd helper macros 10(9 minutes ago, 2 files, 6+ 6-) 13https://github.com/crawl/crawl/commit/c08ff2701284 22:21:43 I try to setup a scenario with 3 actors (player, peaceful orc, and invisible bat) with the existing MockPlayerYouTestsFixture. But that one doesn't reset the env, which means the monsters don't get resetted ... 22:22:41 Midn8 (L15 TeAE) Crash caused by signal #6: Aborted (D:14) 22:23:42 the good thing about that fixture is that it can clean up the save files. should add that to the lua fixture. 22:25:08 i guess i will just use the lua fixture to write test in my fork, and just submit documentations. 22:25:14 vortex (L22 DEFE) ERROR in 'prompt.cc' at line 348: ASSERT failed: the_index of 52 out of range 0 (0) .. ENDOFPACK (52) (Elf:2) 22:26:01 vortex that guy again? I guess it is race condition between dlua and clua because he uses coroutine in clua. 22:27:03 wouldn't he learn that there are dark corners in the world that he shouldn't look at, go to, or even think about?? 22:27:38 vortex (L22 DEFE) ERROR in 'prompt.cc' at line 348: ASSERT failed: the_index of 52 out of range 0 (0) .. ENDOFPACK (52) (Elf:2) 22:28:45 vortex (L22 DEFE) ERROR in 'prompt.cc' at line 348: ASSERT failed: the_index of 52 out of range 0 (0) .. ENDOFPACK (52) (Elf:2) 22:29:09 wait. now even the crash handler get dragged into the mess?