00:00:10 03Skrybe02 {ebering} 07* 0.25-a0-820-g0199dc3: Update layout_cellular.des 10(5 weeks ago, 1 file, 6+ 5-) 13https://github.com/crawl/crawl/commit/0199dc354092 00:02:59 Unstable branch on crawl.kelbi.org updated to: 0.25-a0-820-g0199dc3540 (34) 00:36:58 03Peter Gerlagh02 {ebering} 07* 0.25-a0-821-ge3a6afd: Remove autotravel water disincentive for merfolk 10(12 days ago, 1 file, 2+ 5-) 13https://github.com/crawl/crawl/commit/e3a6afd762a0 00:38:34 03riverfiend02 {ebering} 07* 0.25-a0-822-g6581751: Add the possibility of one more Xom demon 10(5 days ago, 1 file, 1+ 1-) 13https://github.com/crawl/crawl/commit/6581751b113f 00:48:12 Unstable branch on crawl.kelbi.org updated to: 0.25-a0-822-g6581751b11 (34) 02:13:10 !seen reaverb 02:13:10 I last saw reaverb at Sat Feb 15 06:49:26 2020 UTC (9w 1d 23h 23m 43s ago) quitting, saying 'Quit: Leaving'. 02:13:38 !tell reaverb Your catch2 species tests just picked up a particularly easy to overlook error in some code I was writing -- thank you! 02:13:39 alexjurkiewicz: OK, I'll let reaverb know. 02:45:03 Monster database of master branch on crawl.develz.org updated to: 0.25-a0-651-g8c222c7a2b 03:24:50 Unstable branch on crawl.beRotato.org updated to: 0.25-a0-822-g6581751 (34) 05:21:26 03Alex Jurkiewicz02 07https://github.com/crawl/crawl/pull/1368 * 0.25-a0-823-ga90bf34: Add util/tag-35-upgrade.py 10(3 minutes ago, 1 file, 387+ 0-) 13https://github.com/crawl/crawl/commit/a90bf34b1a05 05:21:26 03Alex Jurkiewicz02 07https://github.com/crawl/crawl/pull/1368 * 0.25-a0-824-g2ac55d8: Fix compilation with TAG_MAJOR_VERSION 35 10(3 minutes ago, 20 files, 83+ 39-) 13https://github.com/crawl/crawl/commit/2ac55d87c5b2 05:21:26 03Alex Jurkiewicz02 07https://github.com/crawl/crawl/pull/1368 * 0.25-a0-825-g311d087: Add CI jobs to test TAG_MAJOR_VERSION 35 10(3 minutes ago, 1 file, 32+ 4-) 13https://github.com/crawl/crawl/commit/311d0870a522 05:59:40 advil: can you elaborate on pkg-config bad on macos? 08:25:50 03Aidan Holm02 07[trap-features] * 0.25-a0-823-ge86166c: Rename trap category -> trap feature 10(4 days ago, 6 files, 14+ 15-) 13https://github.com/crawl/crawl/commit/e86166c5235f 08:25:50 03Aidan Holm02 07[trap-features] * 0.25-a0-824-gc5703f2: Add individual feature types for mechanical traps (#12170) 10(4 days ago, 8 files, 98+ 3-) 13https://github.com/crawl/crawl/commit/c5703f2676ff 08:25:50 03Aidan Holm02 07[trap-features] * 0.25-a0-825-g30a70f5: Split teleport trap feature in two (#12170) 10(4 days ago, 6 files, 12+ 3-) 13https://github.com/crawl/crawl/commit/30a70f546054 08:31:34 alexjurkiewicz: was there any sort of dashboard set up for code coverage? or is it just a makefile rule for now? 09:04:39 oh, nvm, found codecov.io 09:18:30 alexjurkiewicz it's not bad it's just not reliable ime 09:19:33 (it's also not very mac-y) 09:33:19 03Aidan Holm02 07* 0.25-a0-823-ge86166c: Rename trap category -> trap feature 10(4 days ago, 6 files, 14+ 15-) 13https://github.com/crawl/crawl/commit/e86166c5235f 09:33:19 03Aidan Holm02 07* 0.25-a0-824-gc5703f2: Add individual feature types for mechanical traps (#12170) 10(4 days ago, 8 files, 98+ 3-) 13https://github.com/crawl/crawl/commit/c5703f2676ff 09:33:19 03Aidan Holm02 07* 0.25-a0-825-g30a70f5: Split teleport trap feature in two (#12170) 10(4 days ago, 6 files, 12+ 3-) 13https://github.com/crawl/crawl/commit/30a70f546054 09:37:24 advil: apart from the macos pkg-config thing, any other objections to pull/1367? 09:37:54 I didn't look at most of the commits but if it passes tests it's probably fine 09:38:05 I'm planning on handling the pkg-config thing differently in another commit btw 09:38:25 since the issue that PR raises is how would a distribution that wants to use it enable it 09:39:06 i was thinking about a util/pkg-config wrapper script 09:39:08 honestly I think it's not worth putting that much effort into if the cmake thing is going to go anywhere 09:39:22 I was just going to add a variable you could use to force it on the cl 09:39:48 + documentation (which isn't there right now) 09:41:11 I actually think making a wrapper script would be much better and much less painful than "just one more" makefile variable 09:41:26 for starters, you'd have a proper programming language 09:41:28 I have occasionally noticed that we have different tastes in pain 09:41:33 heh 09:41:41 ls util/ | wc -l 09:41:41 78 09:42:05 mind you I wouldn't swear we don't have 78 misc capitalized Makefile vars 09:43:50 hm, a quick hacky pipeline: rg -o '[A-Z_]+' Makefile | sort | uniq | rg '...' | wc -l 09:43:53 309 09:45:36 that's actually guite a bit more than I expected 10:09:43 well, that is a lot 10:10:04 03Emily02 {Aidan Holm} 07* 0.25-a0-826-gc0736c7: Use pkg-config for SQLite 10(2 days ago, 1 file, 10+ 29-) 13https://github.com/crawl/crawl/commit/c0736c7dca71 10:10:04 03Emily02 {Aidan Holm} 07* 0.25-a0-827-gdbf7afb: Drop support for ANCIENT_SQLITE 10(2 days ago, 1 file, 0+ 5-) 13https://github.com/crawl/crawl/commit/dbf7afbd95ce 10:11:03 that is picking up a lot of stuff like GCC 10:11:57 Unstable branch on crawl.kelbi.org updated to: 0.25-a0-825-g30a70f5460 (34) 10:12:34 hey, I never said it was accurate :D 10:15:17 I do feel like util/ is a bit of a wilderness of misc scripts with varying levels of attention, whereas the Makefile at least gets regular views 10:16:13 then i'd say the proper solution is to add linting and CI checks for those scripts 10:16:20 (and get rid of the perl...) 10:17:47 Unstable branch on crawl.kelbi.org updated to: 0.25-a0-827-gdbf7afbd95 (34) 10:18:12 and cleanly separate util/ into dev utilities and scripts that are part of the build 10:19:27 right, the problem is partly that it has a bit of a `contrib` status 10:19:39 although I actually don't mind all the gen-foo programs; it's very clean and redo-like 10:20:20 if you're going to go whole hog with code gen, this is probably one of the better ways to do it 10:59:29 what is a polite way to say, "crawl builds on linux so maybe do a bit more work than just pasting a build error from an automated system into a github issue" 11:04:08 toughie 11:04:19 looking at their full build log it looks like they're applying custom patches!? 11:05:05 + /usr/bin/patch --no-backup-if-mismatch -p1 -b --suffix .crawl_bin --fuzz=0 11:05:06 patching file crawl-ref/source/Makefile 11:05:14 I'd uh, close the issue, 11:05:21 we support our build 11:06:12 well, it's possible that this is like the last one which was basically using some build argument that is in the Makefile but not well tested 11:06:21 but I posted a set of questions that encourages them to do a bit more work 11:33:49 is there any mailing list for announcing breaking compatibility with makefile/build stuff? 11:34:28 you mean something that would affect like packaging of this sort? 11:34:34 not that I know of 11:34:37 yeah 11:35:33 we don't have a lot of contact with the people doing this on the whole 11:35:55 except that 1kb still does the debian packaging so we *definitely* hear about that if he gets annoyed at something 11:38:09 could be worse 11:43:36 at one point last year I did a pass through a few distributions' bug trackers and fixed some issues there that had never been reported 11:44:28 so while I think maybe this reported is underestimating how much they've deviated from the supported build process (and any help we offer is kind of pro bono), it's still useful in a way to see what distributions need to solve by patching 11:44:35 *this reporter 11:45:54 03Goratrix02 {Aidan Holm} 07* 0.25-a0-828-ga6db794: Add unused unidentified mechanical trap tile (#12170) 10(4 weeks ago, 1 file, 0+ 0-) 13https://github.com/crawl/crawl/commit/a6db7940a56f 11:46:24 aidanh do you have any immediate guesses why iswalnum might not be available under some specific build circumstances? If you don't know offhand it's not obviously worth spending a lot of effort on 11:46:49 no clue; i copied that code in from the cio.cc stuff iirc 11:47:14 it might be something about it being used in a specific namespace in combination with these random hardening things that redhat seems to do 11:47:24 if they can reproduce that with that huge gcc invocation, it should be easy enough to narrow down 11:56:12 example, this report from fedora, even though I couldn't replicate the crash (https://bugzilla.redhat.com/show_bug.cgi?id=1575324) led to this fix for a potential segfault, where the code was indeed wrong (https://github.com/crawl/crawl/commit/0c774a7e83d7) 11:56:22 so I basically want to encourage them 11:57:07 but in this case the answer may need to be, "your build is really, really different from anything we support, please give us a MWE that replicates the build failure" 12:01:38 03Aidan Holm02 07* 0.25-a0-829-g4e6a552: Fix missing tile in popup description 10(45 seconds ago, 1 file, 2+ 0-) 13https://github.com/crawl/crawl/commit/4e6a552925da 12:02:49 Unstable branch on crawl.kelbi.org updated to: 0.25-a0-828-ga6db7940a5 (34) 12:03:46 that sort of thing is definitely nice 12:04:13 i'm just concerned by the paralysis caused by never knowing what's actually being used by someone, and the ensuing inability to refactor 12:04:51 my take is that these maintainers are generally able and willing to do the minor patching needed when something explicitly changes, especially if we've documented it 12:05:16 and they all have automated processes that flag new releases, and also build failures 12:05:39 so I basically don't think about them when improving the Makefile 12:06:26 so for example if you're worried about package maintainers with the font param the you mentioned in #1367, just go ahead and do it and they'll adapt 12:12:06 that, less so; it's more the reams of tangled makefile crap 12:12:20 although that font one does seem relatively low-risk, to be sure 12:12:51 ugh, I guess I should give the makefile a break and get back to widget-ifying the main crawl game 12:17:46 yeah, not sure it's a particularly rewarding thing to work on and to a large extent it basically does Work, it's just that it could be a lot better 12:18:16 Unstable branch on crawl.kelbi.org updated to: 0.25-a0-829-g4e6a552925 (34) 12:18:38 so it's again one of these things at a local maxima where messing with it is as likely to knock it off that maxima as much as anything 12:19:16 I haven't been putting this into practice but personally when I think about how I want to expend effort on the build system, I think it would be better aimed at cmake 13:31:53 Unstable branch on crawl.akrasiac.org updated to: 0.25-a0-829-g4e6a552 (34) 14:41:49 yeah, cmake is probably a better bet, although I can't help but feel it's going to descend into the same sort of tarpit 14:42:13 all this IF STREQUAL stuff 14:43:57 i have a hacked up redoconf build setup working, although that might be a bit too revolutionary 14:44:25 although apparently one of linley's new games uses redo, at least 15:02:29 not sure I know that one 15:55:43 -!- werekitten is now known as misha 17:38:50 03Alex Jurkiewicz02 07https://github.com/crawl/crawl/pull/1368 * 0.25-a0-826-g3c7bfd2: fixup! Add util/tag-35-upgrade.py 10(50 seconds ago, 1 file, 10+ 10-) 13https://github.com/crawl/crawl/commit/3c7bfd261366 17:50:22 the mechanical trap changes that were merged overnight break TAG 35 compatibility :) 17:50:53 feature-data.h:257:6: error: ‘DNGN_TRAP_MECHANICAL’ was not declared in this scope 17:54:50 ah, easy fix, i'll add to my PR 18:23:21 Unstable branch on underhound.eu updated to: 0.25-a0-829-g4e6a552925 (34) 18:28:24 pretty much all of that code doesn't get tested unless it introduces a minor version 21:13:24 yeah I didn't bother with a minor version in this instance 21:16:44 I guess if we do get CI working for major version bump, we'll be testing it from now on 21:37:22 most code should at least add some ifdefs and ifdefs though 21:37:41 mine did, but I guess I forgot a case or two 21:38:04 yeah, CI for that will be great 23:17:29 ugh, I've been updating the comments in rltiles and testing my TAG upgrade script. But every now and then I forget to commit my intermediate changes and lose 10mins of work :\ 23:28:02 because you run the script and it thus removes the tagging? yeah, hard to fix that with anything other than forcing yourself to make commits, I guess 23:28:56 you could have another repo that pulls the commits and only run your tests from the second repo 23:29:14 yeah, exactly. I've nearly finished now. Doing it once for all my changes learned me quick