03:34:22 Experimental (bcrawl) branch on underhound.eu updated to: 0.23-a0-5261-gd9800d219b 05:17:04 Unstable branch on crawl.akrasiac.org updated to: 0.35-a0-389-gd87ce67 (34) 07:37:54 <05icoson> apparently someone vibe coded a mobile webtiles client 07:41:54 <06pleasingfungus> how nice 07:42:33 <05icoson> the code actually doesn't look completely crazy to me 07:42:48 <05icoson> slightly odd how extremely anonymous whoever did this is being 07:43:25 <05icoson> https://pocketzot.pages.dev/about 07:44:57 <06pleasingfungus> funny to me to beg donations for ai-written code 07:54:07 <08nicolae> those tokens aren't free 08:46:38 I would assume it's a continuation of https://github.com/eaudetcobello/webtiles-pwa given that was also hosted on pages.dev (see also https://github.com/crawl/crawl/pull/5255) 09:56:38 03CrawlOdds02 {DracoOmega} 07* 0.35-a0-390-g4c4374f0c3: Simplify piety gain 10(5 months ago, 2 files, 130+ 216-) 13https://github.com/crawl/crawl/commit/4c4374f0c3b4 09:56:38 03CrawlOdds02 {DracoOmega} 07* 0.35-a0-391-g69972881b1: Remove piety decay 10(5 months ago, 3 files, 32+ 78-) 13https://github.com/crawl/crawl/commit/69972881b14d 09:56:38 03CrawlOdds02 {DracoOmega} 07* 0.35-a0-392-g73cf9495f3: Fix TSO piety bug and rebalance 10(5 months ago, 1 file, 7+ 9-) 13https://github.com/crawl/crawl/commit/73cf9495f334 09:56:38 03CrawlOdds02 {DracoOmega} 07* 0.35-a0-393-g55c1f3f727: Remove piety hysteresis 10(5 months ago, 4 files, 7+ 29-) 13https://github.com/crawl/crawl/commit/55c1f3f72735 09:56:38 03CrawlOdds02 {DracoOmega} 07* 0.35-a0-394-g89d6a61940: Make sure we excommunicate on piety 0 10(5 months ago, 1 file, 5+ 3-) 13https://github.com/crawl/crawl/commit/89d6a6194095 09:56:38 03CrawlOdds02 {DracoOmega} 07* 0.35-a0-395-gfa4c3daf6d: Make Okawaru piety a bit less random. 10(3 months ago, 3 files, 4+ 5-) 13https://github.com/crawl/crawl/commit/fa4c3daf6d97 09:56:38 03DracoOmega02 07* 0.35-a0-396-g8a323bdabc: Remove a mildly confusing ancient comment 10(2 weeks ago, 1 file, 0+ 3-) 13https://github.com/crawl/crawl/commit/8a323bdabcd9 09:56:38 03DracoOmega02 07* 0.35-a0-397-g00792b2873: Remove some code preventing djinni from getting Sif/Veh/Kiku gifts 10(2 weeks ago, 3 files, 1+ 20-) 13https://github.com/crawl/crawl/commit/00792b2873c0 09:56:38 03DracoOmega02 07* 0.35-a0-398-gc4030cd8d8: (Mostly) eliminate the effect of god gifts on piety 10(2 weeks ago, 2 files, 6+ 14-) 13https://github.com/crawl/crawl/commit/c4030cd8d858 09:56:38 03DracoOmega02 07* 0.35-a0-399-g491ad77c2f: Tweak Brothers in Arms 10(13 days ago, 8 files, 75+ 108-) 13https://github.com/crawl/crawl/commit/491ad77c2fb5 09:56:38 ... and 13 more commits 10:07:16 03hellmonk02 07* 0.35-a0-413-g1d5e417ec5: fix another docs typo 10(71 seconds ago, 1 file, 3+ 3-) 13https://github.com/crawl/crawl/commit/1d5e417ec5c9 11:05:52 03CrawlOdds02 {GitHub} 07* 0.35-a0-414-g31e5955297: Give phantom mirror a custom targetter (#5267) 10(61 seconds ago, 4 files, 35+ 5-) 13https://github.com/crawl/crawl/commit/31e5955297dd 11:33:31 03CrawlOdds02 {GitHub} 07* 0.35-a0-415-g2b0c0ba595: Adjust Tempering autotargeting (#5266) 10(19 seconds ago, 1 file, 1+ 1-) 13https://github.com/crawl/crawl/commit/2b0c0ba595ee 11:52:41 03DracoOmega02 07* 0.35-a0-416-ga199ff0106: Properly give Sif titles when Spellcasting is the highest skill (Ge0FF) 10(3 minutes ago, 1 file, 2+ 0-) 13https://github.com/crawl/crawl/commit/a199ff0106c8 11:54:49 03Sergey Semushin02 {GitHub} 07* 0.35-a0-417-g4c63488e59: Fix Antimagic Gaze description (#5262) 10(5 seconds ago, 1 file, 1+ 2-) 13https://github.com/crawl/crawl/commit/4c63488e5966 11:58:01 03hellmonk02 07* 0.35-a0-418-gef280457d8: update credits 10(58 seconds ago, 1 file, 1+ 0-) 13https://github.com/crawl/crawl/commit/ef280457d822 12:16:16 04Build failed for 08master @ 4c63488e 06https://github.com/crawl/crawl/actions/runs/26532036175 12:19:50 I'm assuming this is a wizmode thing, but when fully transcendent in all skills I'm getting "You cannot train any new skills!" every single time I gain XP 12:51:18 <04dracoomega> That is really supposed to only happen once 12:51:31 <04dracoomega> C++ if (!you.received_noskill_warning) { you.received_noskill_warning = true; mpr("You cannot train any new skills!"); } 12:51:55 yeah I'm not sure how it's happening either. The only time it should be unset is when the skill menu is opened (and opening and closing it doesn't fix it) 13:26:00 hey @wizardike, I'm looking at some of the curses input handling stuff specifically regarding ^H versus ASCII DEL. On macOS, it seems that DEL is getting reported by curses (either bundled curses or a modern version) as not a key code but as something that's printable, meaning libunix.cc:678 needs to be case -127: instead of case 127: to function correctly. I have to assume this is somehow a macOS issue because (unless SSH translates DEL to 13:26:00 ^H which is probably possible) I don't see the issue on a linux box. 13:27:08 relevant links: https://github.com/crawl/crawl/blob/9e2179861ec4accfd3ced83639dcd004ba15abcb/crawl-ref/source/libunix.cc#L678 and https://github.com/crawl/crawl/blob/9e2179861ec4accfd3ced83639dcd004ba15abcb/crawl-ref/source/libunix.cc#L528-L547 13:29:50 Also—this issue is only noticeable in iTerm2 for me and not in Terminal.app as the latter by default translates backspace to ESC[3~ while the former doesn't have a similar setting (one that maps backspace to ^H) enabled by default 13:34:27 (Also as another weird issue regarding that second linked bit of code—for some reason I need to change "curses_pending_key = ESCAPE" to something else to attach a debugger because whenever the debugger attaches get_wch reports an error for some reason which crawl then interprets as an escape, immediately quitting it) 14:52:40 <04dracoomega> Speaking of negative keycodes, I was looking into https://github.com/crawl/crawl/issues/5273 and the issue seems to be a crash from calling isspace() on a numpad keycode (which is negative). It doesn't crash for me in local tiles, but does in webtiles. But when I went to check how long this issue might have been around by using CDI, it doesn't cash there in any version. But does crash on CNC (which is where the player who 14:52:41 reported it appears to play). Supposedly, calling isspace() on negatives is undefined behavior (and may also be locale-specific? Which could explain CNC crashing and not CDI). Casting the keycode to an unsigned char does fix the crash for me locally, but I have some concerns that this isn't entirely the correct solution. Given that such casts merely discard the higher-order bytes, is it possible there could be some collision with space? (ie: we don't 14:52:41 know for sure that the keycode even is a character, let alone whether that character is a space or not). I don't suppose anyone with more relevant technical chops might be able to weigh in? (The code in question can be found in invent.cc:1981) 14:54:17 Yeah was similarly dubious of such a fix for my former issue. Also, dear goodness?? Someone should make it so computers aren't so complicated. 14:56:53 <04dracoomega> There's an irony that pressing space doesn't even reach that part of the code by default anyway 14:57:10 <04dracoomega> Since above that there's a C++ else if (key_is_escape(keyin) || allow_easy_quit && keyin == ' ') { ret = PROMPT_ABORT; break; } 14:58:13 <04dracoomega> (Though this is disabled in a couple menus: a god blessing a weapon, and Ash cursing/uncursing something. Apparently?) 14:58:47 <04dracoomega> Not entirely sure why 14:59:20 <04dracoomega> (Like, is there some reason why it is bad to allow spacebar to exit those specific menus but not any other one??) 15:03:27 <04dracoomega> Yeah, space also just does nothing there. No messaging or cancelling or anything. (Worth noting: you don't even get to the part of that menu that even runs this code unless you first press a letter corresponding to an invalid item, which closes the initial inventory popup and enters a different message loop) 15:04:13 <04dracoomega> So is this special-case handling to ensure that in these couple of menus, if you first pick something invalid and then press space, it doesn't go "Huh?" Is that what this isspace() is actually doing? 15:04:14 <04dracoomega> ??? 15:05:39 dracoomega in a few hours: https://www.youtube.com/watch?v=HFgeustBpFk 15:05:41 <04dracoomega> So I think I may be able to fix this issue by just removing this function call entirely 15:06:09 <04dracoomega> And we can just 'Huh?' at any input at all at this point 15:35:08 03DracoOmega02 07* 0.35-a0-419-g033806e58b: Fix a crash-on-demand in some webtiles versions, remove a menu param 10(5 minutes ago, 3 files, 6+ 15-) 13https://github.com/crawl/crawl/commit/033806e58ba2 15:42:53 Unstable branch on underhound.eu updated to: 0.35-a0-418-gef280457d8 (34) 17:49:12 03WizardIke02 07* 0.35-a0-420-g7caaed802a: Fix backspace sometimes not working on macOS (perryprog) 10(13 minutes ago, 1 file, 5+ 3-) 13https://github.com/crawl/crawl/commit/7caaed802a24 17:50:03 Huh fair enough 🫡 17:55:09 I think that'll also apply to freebsd, although I don't know how many freebsd users you might have 17:55:19 (that's where macos inherited it from) 17:57:07 Yeah I was assuming that other platforms were sending it with the sign flipped but I had no way to test that so /shrug 17:58:06 I don't think they do 17:58:24 can't even blame signed vs. unsigned char because 127 is in range for both 18:01:06 I mean I think it'd be if curses reports it as a keycode (via get_wch returning a KEY_CODE_YES) or not and yeah I suppose there's no reason that'd vary across platforms. The comment for the other case (line 541) saying "a normal (printable) key" is what really through me off though since I think that's not really true 18:01:10 hm. negative is synthetic keys. I suspect someone confused DEL and KEY_DC; checking blame 18:01:36 you're about to find out why I picked whom I pinged :P 18:01:51 depends on your definition of "printable"; it'd go with what I just said if they're distinguishing from ncurses synthetic keys 18:02:06 yeah I think you're right 18:04:10 oh lol 18:05:12 2006 it was correct and even had a comment for it https://github.com/crawl/crawl/blame/d412b83276b10a1d652ed3ed36eccda4610ec31b/crawl-ref/source/libunix.cc#L578 18:05:32 the next commit, which merged two almost-identical get-character functions, broke it 18:05:47 oh! Bizarre! 18:07:40 and the comment on it does in fact reflect confusion between KEY_DC and DEL 18:07:58 someone thought they were the same thing, but DEL is standard ASCII 18:08:23 (and, practically, del is delete-backwards where DC is delete-forwards) 18:08:45 and backspace, at least originally, backspaced but didn't delete 18:09:03 the move from keypunches to terminals changed that 18:10:38 ("delete" on a keypunch punched all the holes; card readers treated that as a space) 22:35:59 Unstable branch on crawl.develz.org updated to: 0.35-a0-420-g7caaed802a (34) 22:59:47 Windows builds of master branch on crawl.develz.org updated to: 0.35-a0-420-g7caaed802a 23:35:07 Unstable branch on cbro.berotato.org updated to: 0.35-a0-420-g7caaed802a (34) 23:56:09 Monster database of master branch on crawl.develz.org updated to: 0.35-a0-420-g7caaed802a