00:13:39 Unstable branch on CRAWL.XTAHUA.COM updated to: 0.24-a0-420-g24607be9ce (34) 00:15:03 Stable (0.23) branch on CRAWL.XTAHUA.COM updated to: 0.23.1-79-g0e89c35649 00:31:35 03crystalsolenoid02 07https://github.com/crawl/crawl/pull/1080 * 0.24-a0-430-gc6c0214: directn.cc conjugates mostly correctly for "they" 10(5 minutes ago, 1 file, 38+ 19-) 13https://github.com/crawl/crawl/commit/c6c0214f84c3 01:20:51 Experimental (bcrawl) branch on underhound.eu updated to: 0.23-a0-1915-g10a67aac0f 01:41:35 03crystalsolenoid02 07https://github.com/crawl/crawl/pull/1080 * 0.24-a0-431-g398f6c6: Fixed unbrace requirement 10(2 minutes ago, 1 file, 18+ 0-) 13https://github.com/crawl/crawl/commit/398f6c65a6ba 03:12:05 Unstable branch on crawl.beRotato.org updated to: 0.24-a0-420-g24607be (34) 03:30:50 Fork (bcrawl) on crawl.kelbi.org updated to: 0.23-a0-1915-g10a67aac0f 04:39:30 -!- amalloy is now known as amalloy_ 05:35:28 -!- bh is now known as Guest31949 07:31:28 Vehumet enhances range of evocables, part II. 13https://crawl.develz.org/mantis/view.php?id=12007 by Yermak 09:23:05 Stable (0.22) branch on underhound.eu updated to: 0.22.1-59-ge965b3aba3 11:22:50 Stable (0.23) branch on underhound.eu updated to: 0.23.1-79-g0e89c35649 13:05:43 Unstable branch on crawl.akrasiac.org updated to: 0.24-a0-420-g24607be (34) 13:42:14 Guys, I tried setting a reminder when AK enters Gauntlet, but it doesn't work: https://pastebin.com/5Z6k3jJr Could you point out why? 13:45:02 -!- amalloy_ is now known as amalloy 13:56:22 i'm far from an rcfile expert, but that looks like it's at the top level of your rcfile, so it's only evaluated when you start/load a game. you'd want something similar in your ready() function instead 14:01:19 Thank you, it worked! 14:31:35 03MarvinPA02 07* 0.24-a0-421-g6863ae8: Don't allow Vehumet to extend lightning rod range (#12007) 10(23 minutes ago, 1 file, 2+ 1-) 13https://github.com/crawl/crawl/commit/6863ae89ae89 14:47:25 Unstable branch on crawl.kelbi.org updated to: 0.24-a0-421-g6863ae89ae (34) 16:09:43 here's a very crawlcode question: if there is a label on a for loop, and a goto call for that label from inside the loop, does it restart the loop? 16:10:26 or an 80s C question I suppose 16:10:47 Yes 16:52:31 03advil02 07* 0.24-a0-422-g9eff389: Optimize find_maps_for_tag somewhat 10(9 minutes ago, 1 file, 2+ 1-) 13https://github.com/crawl/crawl/commit/9eff3895ba7d 16:53:28 if anyone had the impression that overflow vault selection had gotten slower lately, well, you were right 17:00:47 !source ng-init.cc:273 17:00:48 https://github.com/crawl/crawl/blob/master/crawl-ref/source/ng-init.cc#L273 17:02:26 so can someone explain to me under what conditions this bit of code terminates? I guess it is when it gets enough continues to get i > size without going all the way through? 17:02:35 Unstable branch on crawl.kelbi.org updated to: 0.24-a0-422-g9eff3895ba (34) 17:03:08 maybe more important, what is it trying to do by doing things this way 17:04:46 It finds some overflow gods to use, throws them away from the unplaced list, starts the loop again. 17:06:44 specifically, the iterator variable is being used to iterate over all subset combinations 17:07:19 the goto could be refactored as an exterior while (overflow_gods.size() > 0) 17:07:25 Eventually there's none left, i <= size, and the loop body doesn't execute 17:07:57 well replace the goto with a break out of the for loop 17:08:12 and wrap it an extieror loop that does the while(overflow_gods.size() > 0) 17:08:15 actually 17:08:43 03Ozaq02 07https://github.com/crawl/crawl/pull/1059 * 0.24-a0-423-gc9fa2eb: Add Experimental CMake support 10(7 months ago, 8 files, 1704+ 0-) 13https://github.com/crawl/crawl/commit/c9fa2eb8b439 17:10:24 ok, a more specific question: why would you do it this when when there are only 13 temple_overflow_2 vaults and 4 temple_overflow_3 vaults 17:10:29 *this way 17:11:01 do all of the overflow temples support all gods? 17:11:41 (I'm not super familiar, I'm trying to reverse engineer along with you) 17:11:59 no, in fact I don't see any that support anything other than 2 or 3 specific gods 17:12:51 there are generic overflow vaults that are handled separately 17:13:32 I see what you're getting at: it would be much quicker to iterate over the 17 available vaults and check if they provide gods on the target list (based on tag) 17:13:43 yeah 17:14:37 than to iterate over the possibly 2^19 combinations (even if you do break out the first time you get one) when a small temple was chosen 17:15:34 in fairness, I only noticed this because my change a few months ago to represent tags as set made this for loop go a lot slower 17:16:45 that representation probably makes reversing the search (try to pick some specific overflows and see if they match) easier 17:16:49 to write at least 17:23:51 I'd still have to understand this code better to rewrite it 17:47:55 huh I'm not sure anything even handles temple_overflow_1 17:48:59 well, not explicitly, I guess this last loop can probably cause them to place 17:57:36 advil: did you still need those seed tests, btw? 17:57:45 or that specific seed test you asked for, rather 17:58:07 and did you need that specific and exact order, D:1-15, Temple, Lair:1-5 I think it was 18:02:05 yeah, specific exact order 18:02:18 it's the pregen order up to the point where I think your game diverges 18:02:49 so that I can compare the builder output from lair 5 18:03:48 let me try it 18:08:11 -!- Tiobot is now known as Guest83973 18:10:54 advil: This should work; it looked good at game start and I saw slime temple after going D:1-15, and I then went Lair:1-5: https://drive.google.com/file/d/1mn6PLTW1bCB-ti_6ea1bP_bPQiehlwUd/view?usp=sharing 18:22:16 Unstable branch on underhound.eu updated to: 0.24-a0-422-g9eff3895ba (34) 18:51:43 Krim260 (L27 GnWn) Crash caused by signal #15: Terminated (Zot:5) 18:53:08 I just got shafted while changing armour. 18:53:34 It seems that trap activation sometimes (always?) happens a turn later. 18:55:30 Crawl turn system really needs to be overhauled. 18:58:02 Other things that happen a turn later: Slow debuff from elephant slug (appears later, disappears later); Usk paralysis; orb spiders after being EHed move backwards. 19:03:40 Trap activation isn't in the same code sequence point as the other things you point out 19:04:10 It happens during world reacts, which happens every turn (with the exception of post-pastwall, where you get an action at the endpoint before the world reacts) 19:04:51 Were you changing armour either 1) after paswalling 2) near unexplored territory around fog 3) near a closed door with the other side unexplored? 19:43:51 ebering, right after the teleport 19:53:28 -!- amalloy is now known as amalloy_ 22:25:09 -!- amalloy_ is now known as amalloy 22:25:24 Yermak: just to check, is this in trunk, or 0.23? 22:27:20 gammafunk: thanks, but that's a debug-lite build isn't it :) 22:27:54 yes, but the build I saw divergence with was debug-lite 22:27:56 you need full debug? 22:30:52 advil: ^ 22:31:14 yeah, if possible 22:31:24 what I need to look at is in the debug msg spam 22:31:26 ok, can do 22:31:59 hrm, guess I didn't realize that a save made in e.g. debug lite was different 22:32:04 it's the log itself that I was hoping for 22:32:16 oh I see, so the log that's going to be generated as I traverse 22:32:19 yeah 22:32:21 gotcha 22:33:42 unrelated, but on my system I now get the following compilation warning 22:33:51 HOSTCXX tool/tile_page.o 22:33:53 tool/tile.cc: In member function ‘void tile::copy(const tile&)’: 22:33:56 tool/tile.cc:295:76: warning: ‘void* memcpy(void*, const void*, size_t)’ writing to an object of type ‘class tile_colour’ with no trivial copy-assignment; use copy-assignment or copy-initialization instead [-Wclass-memaccess] 22:34:00 memcpy(m_pixels, img.m_pixels, m_width * m_height * sizeof(tile_colour)); 22:34:03 I think it's a result of the newer gcc on ubuntu 19 22:34:06 as I didn't get this with the gcc on 18 22:34:23 my gcc is 8.3.0 22:37:22 ubuntu 19 is the one thing I haven't actually tried to replicate it, I guess I should still do that at some point 22:41:32 looks like the solution to that warning is to not use memcpy on anything non-trivial 22:44:03 well, I suppose even testing gcc 8 is probably a good idea 22:44:06 advil: here you go: https://drive.google.com/file/d/154V-XM3H9U2gBhyGJb41rSULhGpDO3M-/view?usp=sharing 22:48:08 gammafunk thanks, that has what I need 22:48:14 can at least confirm that the divergence happened there 23:50:20 -!- TigerbotHesh_ is now known as TigerbotHesh