00:04:24 Webtiles server stopped. 00:04:29 Webtiles server started. 00:07:39 <09g​ammafunk> @advil Installed 0.27 on CAO and will continue checking it to make sure I haven't messed anything up. Spent a long time cleaning up the crawl-dev dgamelaunch-config repo to put all recent changes made in nice commits, so that repo is in a better state now. 00:49:28 <08n​icolae> that is a good tile 01:34:32 Unstable branch on crawl.develz.org updated to: 0.28-a0-43-g560ac104fe (34) 01:55:36 Windows builds of master branch on crawl.develz.org updated to: 0.28-a0-43-g560ac104fe 02:16:44 Unstable branch on cbro.berotato.org updated to: 0.28-a0-43-g560ac104fe (34) 02:41:45 03gammafunk02 07* 0.28-a0-44-gf34e66f144: doc: Mention Storm Form in the changelog 10(85 seconds ago, 1 file, 9+ 0-) 13https://github.com/crawl/crawl/commit/f34e66f1441c 02:42:36 03gammafunk02 07[stone_soup-0.27] * 0.27-b1-41-gafb5482f4b: doc: Mention Storm Form in the changelog 10(2 minutes ago, 1 file, 9+ 0-) 13https://github.com/crawl/crawl/commit/afb5482f4bc3 02:46:47 Unstable branch on crawl.kelbi.org updated to: 0.28-a0-44-gf34e66f144 (34) 02:53:49 Monster database of master branch on crawl.develz.org updated to: 0.28-a0-43-g560ac104fe 03:32:16 04Build failed for 08master @ f34e66f1 06https://github.com/crawl/crawl/actions/runs/1073910389 03:34:10 Stable (0.27) branch on cbro.berotato.org updated to: 0.27-b1-41-gafb5482f4b 03:39:39 Stable (0.27) branch on crawl.kelbi.org updated to: 0.27-b1-41-gafb5482f4b 03:40:17 Fork (bcrawl) on crawl.kelbi.org updated to: 0.23-a0-4217-g7c68dc2372 04:17:51 -!- allbery_b is now known as geekosaur 05:03:05 Hunting sling reverting to a state of being unidentifed as shown in this following post: https://pastebin.com/ueMp3uBh 06:58:53 Stable (0.27) branch on underhound.eu updated to: 0.27-b1-41-gafb5482f4b 07:14:47 Webtiles server stopped. 07:14:55 Webtiles server stopped. 07:14:56 Webtiles server started. 07:35:05 railbird80: It goes from the bright green of being equipped to the grey of being unequipped and not an ego item like the flails. Where's the bug? 09:04:30 <06a​dvil> Thanks, sounds good! 09:05:18 <06a​dvil> I'm going to be a bit more around than I expected for the next week but it's because I have shingles, so I think it is a bit of a wash in terms of me doing anything much 10:13:55 <10P​leasingFungus> oh no! 11:06:28 <09g​ammafunk> hope you have a speedy recovery! 11:08:36 <09g​ammafunk> this appeared like some mysterious dark monolith under my message that I could not interpret, at first 11:08:36 <09g​ammafunk> https://cdn.discordapp.com/attachments/747522859361894521/869959533076819998/unknown.png 11:08:44 <09g​ammafunk> dark mode strikes again (in the dark) 11:09:04 <09g​ammafunk> wonder if they've improved light mode to the point where I can live with it now 11:10:00 <09g​ammafunk> gah, it is so very bright. I'll give it a chance though 11:16:03 <10P​leasingFungus> wouldn't recommend it 11:16:07 <10P​leasingFungus> the true dev lurks in darkness 11:16:23 <10P​leasingFungus> it's a slightly more readable message if you agree with it 🙂 11:16:24 <10P​leasingFungus> https://cdn.discordapp.com/attachments/747522859361894521/869961494094315560/unknown.png 11:27:59 <09g​ammafunk> hah 11:28:13 <09g​ammafunk> (I've already switched back to dark mode, I couldn't hack it for even 10 minutes) 11:32:21 <06a​dvil> thanks, I'm 1.5 weeks in so hopefully on the up curve 11:35:21 03advil02 07* 0.28-a0-45-gda2e87d069: fix: off-by-one error in menu calculations 10(51 minutes ago, 1 file, 1+ 1-) 13https://github.com/crawl/crawl/commit/da2e87d06967 11:35:56 03advil02 07[stone_soup-0.27] * 0.27-b1-42-g19c61b19a5: fix: off-by-one error in menu calculations 10(52 minutes ago, 1 file, 1+ 1-) 13https://github.com/crawl/crawl/commit/19c61b19a54e 11:36:54 <09g​ammafunk> interesting dynamic I found after using these commit title prefixes 11:37:19 <09g​ammafunk> for feat and doc you don't feel like you're going to introduce redundant word with the first word or two of the commit title 11:37:36 <09g​ammafunk> but for fix you kind of feel compelled to make it read like "fix: such and such calculation" 11:37:39 <06a​dvil> yeah, it's harder to do imperative fix: descriptions 11:39:11 <09g​ammafunk> I suppose if you do it like you've done and include at least one other word implying a correction without including fix: itself, it's fine 11:41:15 <06a​dvil> man, numpad key handling is such a mess 11:46:06 <06a​dvil> I think I might have finally wrapped my head around one problem, which is that ncurses does not afaict use separate keycodes for numbers on the numpad, but javascript key handling does 11:47:03 Unstable branch on crawl.kelbi.org updated to: 0.28-a0-45-gda2e87d069 (34) 11:47:57 <06a​dvil> and in fact, thought it's entirely possible I'm missing a mangling step, when the javascript keycodes for those keys get sent, ncurses adjusts them into regular number keycodes 11:48:55 <06a​dvil> even though webtiles console does seem to handle those keys differently 11:49:41 ncurses indeed can't tell the difference 11:50:00 <06a​dvil> although I think it may just be that on a mac those keys at console are interpreted not as number keys 11:50:12 there's a not-widely-used numpad mode that causes them to send something like function keys 11:50:34 macs may not support it, like I said it's not widely used 11:50:39 <06a​dvil> well, we are using that, so that's probably what's happening in console (it's not function keys but an adjacent range) 11:50:54 right 11:51:25 <06a​dvil> so possibly a solution would be to manually map numpad keycodes to that range in webtiles 11:53:22 <03w​heals> when i was testing the commit i made the other day numpad numbers seemed to be indistinguishable from normal ones, but maybe that was a failing of the windows console's curses emulation 11:53:36 <03w​heals> i do wonder if I'm the only person running crawl on WSL 11:53:48 <06a​dvil> most windows devices at least have numlock though right? 11:54:06 yes 11:54:09 <03w​heals> indeed 11:54:12 <06a​dvil> numlock usually toggles between that and a mode where they are distinguishable 11:54:29 <06a​dvil> mac does not have numlock (doesn't even have the concept) 11:54:46 well, except on laptops where there's no separate numpad 11:54:52 <06a​dvil> my attention was drawn to this because of someone on a chromebook without numlock though, so it may be becoming more widespread 11:55:19 <06a​dvil> even there there's no numlock, may be fn modifiers 11:56:30 <06a​dvil> an extra layer of annoyance is that the key mapping code in webtiles is not versioned so it's a bit hard to change without breaking who-knows-what 11:56:53 <03w​heals> from what i could tell, numlock switched them from indistinguishable from numbers to indistinguishable from arrows/home/end/etc., but again what I'm doing may have windows-specific twists to curses 11:57:06 <06a​dvil> yeah, I think that's a reasonably normal behavior 11:57:53 <06a​dvil> the usual goal is to be able to bind macros to numbers and use the numpad for navigation, so people with that setup can use numlock off and be ok 11:58:37 <03w​heals> yeah, since arrows would be navigation anyway 11:58:49 <06a​dvil> obv if there is a way to get fully distinct keycodes that allows more flexibility, but I think the normal use cases are ok-ish as long as there is numlock 11:59:46 <03w​heals> i wonder if chromebooks have some software level numlock even if they don't have a button on the keyboard? 12:00:12 <06a​dvil> in the reddit thread people were trying to help the poster find one, but it didn't sound like they'd had luck yet 12:00:55 <06a​dvil> 98% sure even that doesn't exist on mac though 12:01:31 <06a​dvil> really getting rid of numlock does make a lot of sense in 2021 imo, it just doesn't immediately make life easier on us 12:02:23 <03w​heals> i know personally i really like having numlock off when using a word processor and I'm sure there are people who really want it to be on 12:02:36 <03w​heals> I'm just surprised chromebooks even have a numpad really 12:02:46 <06a​dvil> I think this is like the only one that does 12:02:52 <06a​dvil> sounds like the poster bought it for that reason 12:03:00 <06a​dvil> I wonder if there would be a good way to override stuff in key_conversion.js from the versioned code 12:03:43 <06a​dvil> my keyboard just has distinct arrow / home etc keys from the numpad, no need for swapping 😄 12:10:03 <09g​ammafunk> same 12:13:01 <06a​dvil> can someone with a numlock key tell me what keycode shows up in linux (or windows I guess) console, e.g. by seeing how macro entry interprets it, with numlock off/on for keypad 1? 12:13:17 <09g​ammafunk> sure 12:15:15 <09g​ammafunk> @advil with numlock on it's simply 1 and with it off it's \{-1001} 12:15:22 <09g​ammafunk> linux console 12:15:27 huh 12:15:47 that's not what I expected but it still works 12:16:09 <09g​ammafunk> recent ubuntu and in xterm 12:16:29 <09g​ammafunk> still can't get the gnome terminals to properly render certain glyphs 12:16:33 is end a distinct keycode? 12:16:58 <09g​ammafunk> does numlock matter? do you want the dedicated end key or the keypad..oh 12:17:03 <09g​ammafunk> I guess you'd want the dedicated key 12:17:18 <09g​ammafunk> yeah end is 1034 12:17:42 interesting 12:18:04 I have to say I'm always impressed that this stuff usually just works on pro games 12:19:15 on my keyboard 1001 (which we translate to -1001) is numpad 1, and end is something in the 200s, can't remember what exactly 12:19:45 <09g​ammafunk> mine's a fairly recent logitech wireless keyboard 12:19:57 I bet 1034 is what gets used for numpad end, or something like that 12:20:14 <09g​ammafunk> might be interesting to see what my laptop's keyboard does 12:20:17 or what ncurses uses for that 12:20:56 <09g​ammafunk> laptops built-in keyboard behaves identically 12:22:27 <09g​ammafunk> @wheals @wormsofcan if you'd like to join the devil's team again this tournament, feel free to add me as teamcaptain in your RC; I've already added you both to mine, but can remove if you're not joining. hellmonk is also joining us, and if any other dev wants to join, just let me know 12:24:09 <06a​dvil> oh god 12:24:19 <06a​dvil> !source unixcurses_defkeys 12:24:21 <04C​erebot> Can't find unixcurses_defkeys. 12:24:27 <06a​dvil> !source libunix.cc:599 12:24:28 <04C​erebot> https://github.com/crawl/crawl/blob/master/crawl-ref/source/libunix.cc#L599 12:25:19 <06a​dvil> I'm pretty sure I've traveled this exact path before 12:25:41 <06a​dvil> just like last time, I found this function by finding a stack overflow answer here https://stackoverflow.com/questions/16975367/ncurses-reading-numpad-keys-and-escaping 12:29:00 <09g​ammafunk> lol 12:29:07 <09g​ammafunk> that's amazing 12:29:46 <09g​ammafunk> you even commented in that answer 12:35:13 <06a​dvil> I know 12:35:15 <06a​dvil> extremely meta 12:37:28 <06a​dvil> I think I even "fixed" things in this function last time 12:38:45 <06a​dvil> even more amazing, it looks like this was during the last feature freeze 12:39:20 <10P​leasingFungus> time is a flat circle 12:39:21 <06a​dvil> %git b0746d5dc15c 12:39:22 <04C​erebot> advil * 0.26-a0-1201-gb0746d5dc1: Fix macro menu handling of unprintable keycodes (7 months ago, 9 files, 237+ 24-) https://github.com/crawl/crawl/commit/b0746d5dc15c 12:53:15 <06a​dvil> wow, this tool is amazing: https://w3c.github.io/uievents/tools/key-event-viewer.html 12:56:52 <06a​dvil> also, gah, chrome doesn't distinguish keycodes for the numpad at least on my computer 13:01:35 Stable (0.27) branch on crawl.akrasiac.org updated to: 0.27-b1-42-g19c61b1 13:12:29 <09g​ammafunk> I love that we still have these ridiculously convoluted user agent strings even in 2021 13:12:39 <09g​ammafunk> Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) 13:13:19 <09g​ammafunk> it does say Chrome at the end but most of that is not a very meaningful description of the rending engine 13:30:30 Stable (0.26) branch on crawl.akrasiac.org updated to: 0.26.1-11-gd32895e 13:36:40 Unstable branch on crawl.akrasiac.org updated to: 0.28-a0-45-gda2e87d (34) 13:41:19 <06a​dvil> lol /*! jQuery v1.8.2 13:41:24 <06a​dvil> guess I shouldn't be surprised 14:04:47 <10P​leasingFungus> > Sep 20, 2012 14:32:55 <09g​ammafunk> ho boy 14:33:13 <09g​ammafunk> I asked the simple question: "Do wanderers always start with a weapon?" 14:33:26 <09g​ammafunk> And when I began to inspect the wanderer weapon selection code I found: 14:33:55 <09g​ammafunk> cpp skill_type skill = NUM_SKILLS; const skill_type str_weapons[] = { SK_AXES, SK_MACES_FLAILS, SK_BOWS, SK_CROSSBOWS }; int str_size = ARRAYSZ(str_weapons); const skill_type dex_weapons[] = { SK_SHORT_BLADES, SK_LONG_BLADES, SK_STAVES, SK_UNARMED_COMBAT, SK_POLEARMS }; ... switch ((int)role) { case STAT_STR: skill = 14:33:55 _apt_weighted_choice(str_weapons, str_size); break; case STAT_DEX: skill = _apt_weighted_choice(dex_weapons, dex_size); break; ... 14:37:40 <09g​ammafunk> also...is that cast necessary at all? 14:37:49 <09g​ammafunk> it's declared stat_type role 14:45:44 <06a​dvil> can anyone report to me what keycode numpad - gives them in console? 14:51:06 <09g​ammafunk> @advil with numlock: -, without numlock: \{-1018} 14:51:20 <06a​dvil> ok thanks, that's what I get too 14:51:33 <06a​dvil> it's commented as . in libunix.cc but I suspect that's just incorrect 14:54:49 <06a​dvil> I guess if I'm going to break legacy -1010 etc macros that assume those are function keys, on a new version is a good time to do it 14:55:44 <09g​ammafunk> 0.28 will have save compat break (real!) so nothing is off the table as far as breakage goes 15:15:59 <03w​heals> mine does too (which is pretty rare for laptops these days i think), but i still like using the numpad since it puts all the navigation buttons adjacent to each other 15:17:07 <03w​heals> (also mine does - regardless of numlock option but that's consistent with the behavior up till now on WSL) 15:22:14 03advil02 07* 0.28-a0-46-g7b1be97da7: fix: improve handling of numpad keycodes on webtiles 10(20 minutes ago, 8 files, 150+ 60-) 13https://github.com/crawl/crawl/commit/7b1be97da753 15:26:04 Stable (0.27) branch on underhound.eu updated to: 0.27-b1-42-g19c61b19a5 15:29:34 <06a​dvil> hm actions is in a weird state, I'm going to cancel some stuff in the queue 15:34:27 <06a​dvil> we're getting a lot of errors like: Can't find any online and idle self-hosted or hosted runner in the current repository, account/organization that matches the required labels: 'ubuntu-latest' Waiting for a self-hosted or a hosted runner to pickup this job... 15:34:36 <06a​dvil> I wonder if there's some new runner limit in place? 15:41:13 not really sure what's going on, bad timing though 15:43:43 Unstable branch on crawl.kelbi.org updated to: 0.28-a0-46-g7b1be97da7 (34) 15:52:35 <06a​dvil> Test 15:53:11 answer: it prints the thread title in irc but not any messages from threads 15:53:22 oof 15:54:49 <06a​dvil> so yeah, I guess let's not use threads in this channel 15:55:30 <09g​ammafunk> I can probably hack the bridge to pass them through unharmed 15:55:50 <12e​bering> the ui imo leaves things to be desired 15:55:57 -!- allbery_b is now known as geekosaur 15:55:58 <06a​dvil> yeah, it's really odd on desktop client 15:56:06 <06a​dvil> I first tried it on my phone where it seems better 15:57:54 -!- allbery_b is now known as geekosaur 15:58:14 <12e​bering> I said my peace in general since the bridge would probably butcher it 15:58:50 <12e​bering> anyway if we want threaded conversations about crawl dev there's already a great threaded asynchronous messaging platform that's archived and a part of crawl infra 15:58:53 <12e​bering> c-r-d! 16:01:05 <10P​leasingFungus> is it really all that great 16:01:23 <06a​dvil> email kind of sucks for threaded conversation 16:01:59 <10P​leasingFungus> re save compat break: what if we didn't, instead 16:02:35 <06a​dvil> some day we will really run out of minor tags 16:02:42 <09g​ammafunk> we are very close to that, yeah 16:02:49 <06a​dvil> although I do seem to remember that aidanh had managed to implement a workaround for that 16:02:59 <09g​ammafunk> and it's my understanding we can't ...oh 16:02:59 <06a​dvil> and had some dream of rolling save compat 16:03:12 <09g​ammafunk> is there somehow a way we can expand the size of minor tags without a save compat break? 16:03:27 <10P​leasingFungus> intuitively, it seems quite simple to expand the size of major tags 16:03:30 <03w​heals> i thought we already passed the limit before his change 16:03:31 <10P​leasingFungus> and also i thought we'd done so 16:03:38 <06a​dvil> but he's also not-around enough that I don't think he's moved his bouncer to libera chat 16:03:39 <09g​ammafunk> nope 16:03:40 <06a​dvil> "intuitively" 16:03:41 <09g​ammafunk> we're not past that limit 16:03:46 <10P​leasingFungus> i mean, i haven't thought it through in detail 16:03:59 <06a​dvil> the problem is that the header is handled in a somewhat involved way 16:04:11 <06a​dvil> it's not impossible 16:04:18 <06a​dvil> but there's a lot more to work around than it first seems like 16:04:42 <09g​ammafunk> we can only support up to 255 and we're at 234 16:04:58 <10P​leasingFungus> %git 435769570560c5b0e48dffbc07f9cfd4c102d04c 16:04:58 <04C​erebot> Aidan Holm * 0.25-a0-898-g4357695705: Add support for minor tags > 255 (1 year, 3 months ago, 9 files, 147+ 43-) https://github.com/crawl/crawl/commit/435769570560 16:05:03 <10P​leasingFungus> i knew i remembered this 16:05:52 <06a​dvil> oh that did get merged, huh 16:06:39 <10P​leasingFungus> yeah, we're fine 16:06:46 <10P​leasingFungus> yall lust for save compat breakage... smh 16:06:57 <09g​ammafunk> hmmm 16:07:01 <09g​ammafunk> I don't know that we are fine 16:07:17 <10P​leasingFungus> oh? 16:07:58 <09g​ammafunk> this really needs to be well tested with trunks on a few servers 16:08:14 <09g​ammafunk> specifically it's not in use in the wild because we're not at a tag minor above 255 16:08:27 hm? the worst that happens is a save compat break anyway, no? 16:08:45 <09g​ammafunk> well the premise it's we're using this and not breaking saves 16:08:57 <09g​ammafunk> otherwise there's no difference between save compat break or no, right 16:09:34 <10P​leasingFungus> seems like testing is very simple 16:09:41 <10P​leasingFungus> you paste a bunch of tags into tag-version.h 16:09:48 <10P​leasingFungus> you compile 16:10:00 <10P​leasingFungus> you test if you can load the game 16:10:07 <10P​leasingFungus> you check how it looks in an old binary before you added the tags 16:10:09 <06a​dvil> I assume aidanh already did test this 16:10:19 <10P​leasingFungus> also, yes, though it doesn't hurt to double check 16:10:23 <09g​ammafunk> no, I mean it has to be tested on live servers 16:10:25 <10P​leasingFungus> why? 16:10:33 <10P​leasingFungus> what's so complex about this? what sort of strange outcomes do we expect? 16:10:37 <09g​ammafunk> well, because, have you seen how our servers work? 16:10:47 <09g​ammafunk> it definitely should be tested with save uploading, no question 16:11:00 <10P​leasingFungus> sure! 16:11:04 <10P​leasingFungus> sounds like a good project for 0.28 🙂 16:11:05 <06a​dvil> servers should be mostly ok because there's not a lot of opportunity for cross-version loading, and none for forwards-compatibility that I can think of 16:11:05 <09g​ammafunk> also, if we do use this and it works, how do we deal with the continuous code build up? is there a way to do that without save breaking? 16:11:19 <09g​ammafunk> yeah, the save browser issue shouldn't come into effect, right? 16:11:31 <10P​leasingFungus> seems like we could declare "saves before X are no longer supported" 16:11:36 <09g​ammafunk> like the save preview is no issue, that's done using....crawl trunk? 16:11:42 <09g​ammafunk> not actually sure how that part works 16:11:45 <10P​leasingFungus> check for the minor tag while loading and knock out all the relevant code 16:11:54 <06a​dvil> I do think there's the other practical matter which is that ancient save upgrades "work" but at this point there are no guarantees about playability 16:11:58 <10P​leasingFungus> could maintain, say, 5 years of support 16:12:06 <10P​leasingFungus> or some other number 16:12:14 we're already there, I think 16:12:22 <09g​ammafunk> right, so, I mean, this is still in a lot of ways just a save compat break 16:12:23 <10P​leasingFungus> hm? 16:12:26 reoved items, removed monsters 16:12:42 <10P​leasingFungus> sure, but in important ways it's not! 16:12:48 <10P​leasingFungus> you can still continue your character from last year 16:12:52 <10P​leasingFungus> or from a week ago 16:12:53 I had an older game where all the removed monsters showed up as "ghosts" 16:13:10 it had sat for 2-3 years 16:13:10 <06a​dvil> I think this was essentially aidanh's idea for rolling save compat, just he had a way to print informative messages 16:13:11 <10P​leasingFungus> sure 16:13:16 <10P​leasingFungus> seems natural 16:13:54 <09g​ammafunk> that is nice, but my actual concern is not that. it's this: some servers allow save transfers using dgl's system, which (I think?) works with the idea of not transfering a save, but I believe it only knows how to do this using save compat breaking major versions, and other servers force transfer 16:14:12 <09g​ammafunk> it's not just the player happiness but logistically that it doesn't create brokenness for admins 16:14:31 <06a​dvil> that commit preserves backwards save compat 16:14:32 <09g​ammafunk> both dgl and force save transfering have to not render trunk unplayable due to brokenness 16:16:15 <10P​leasingFungus> i don’t understand this sentence, sorry 16:16:34 <10P​leasingFungus> in context, what is a save transfer? 16:17:18 <09g​ammafunk> when a save was created under a different version of crawl compared to the version currently loading it; in the context of online servers, server admins do this in one of two ways 16:17:30 <08w​ormsofcan> re: discord threads, my understanding is you can disable them in the server settings 16:17:46 <08w​ormsofcan> looks like it's an option that can be enabled/disabled per role 16:18:35 <10P​leasingFungus> oh, is the concern that, once we ban loading saves from sufficiently old minor versions, servers will need extra support to avoid transferring those very old saves to a new version? 16:19:47 <10P​leasingFungus> (since the new version will no longer be able to load them) 16:22:50 <09g​ammafunk> yeah, they need to properly deal with that case in terms of not getting stuck in an error loop that would render trunk unplayable, for example. If the dgl scripts think they only have to worry about major version, for example (because that's the concept they were developed around), they might attempt transfer. Or if a server isn't dgl based and always force-transfers, its scripts need to understand 16:22:51 when it can't do that. Presumably they could simply delete the save in the worst-case and start a new character, and the only harm is a lack of prompting. 16:24:39 <10P​leasingFungus> seems like a good reason to be careful w removing support for old minor versions, i agree! 16:24:58 <10P​leasingFungus> but doesn’t seem to present any problem for going beyond 255 minor tags 16:25:58 <09g​ammafunk> My concern about that is only that it actually be tested on a live server to confirm that there is no breakage, because live servers do have the additional scripts that inspect saves etc. Mostly thinking of dgl here. 16:27:45 <06a​dvil> I will say that I'm not sure there any servers in the wild that would need to do upgrades from arbitrarily old versions 16:28:26 <06a​dvil> maybe cdo? but it's configuration is a bit different than the others anyways 16:28:28 <06a​dvil> *its 16:28:52 <09g​ammafunk> yeah, cdo has a lot of custom config, some of which is pretty clever 16:29:28 <06a​dvil> cao at one point would have been an issue but while it was very short on disk space I was force-upgrading trunk games fairly aggressively 16:29:29 <09g​ammafunk> I started mimicing aspects of it for CDI but realized I wanted the test server to be closer to what we have on a "typical" dgl server 16:30:46 <09g​ammafunk> what do you mean by this specifically? all servers have the potential to upgrade from some arbitrarily old trunk version so long as it's post-last-save-compat-break, no? 16:31:07 <06a​dvil> they would need to have such a save around somewhere to upgrade 16:31:23 <06a​dvil> in practice, I don't think any are likely to 16:31:27 <09g​ammafunk> right, around and someone would have to attempt to upgrade it 16:31:37 <06a​dvil> or just load it 16:31:54 <06a​dvil> but that would still lead to the prompt I guess 16:32:03 <09g​ammafunk> but I guess for the purposes of what we're talking about....oh, I guess this is relevant to the discussion of how long we're advertising our compat and want that cutoff is? 16:32:12 <06a​dvil> CAO 16:32:28 <06a​dvil> CAO's window is probably less than a year right now 16:33:12 <09g​ammafunk> right, presumably CXC and CUE could have older saves, but new CXC isn't all that old, and I forget if it copied saves over 16:33:23 <09g​ammafunk> CUE could have some oldish ones, believe it uses dgl 16:34:45 <06a​dvil> possible 16:35:10 <06a​dvil> even CUE might not be >5 years old in its current incarnation 16:37:42 <09g​ammafunk> forget if cwz is older 16:37:52 <09g​ammafunk> !lg * cwz 1 16:37:56 <04C​erebot> 1/1884154. tlatlagkaus the Insei (L1 OpTm), quit the game on D:1 (dpeg_arrival_houses_and_road) on 2015-02-10 03:48:05, with 0 points after 0 turns and 0:00:15. 16:38:28 <09g​ammafunk> cwz has always been webtiles-only and has a force transfer setup, but presumably it could have saves that old in trunk 16:39:01 <09g​ammafunk> of course we don't really have a lot of direct communication with hong and he's always just done whatever 16:41:06 <09g​ammafunk> wow, threads even show up in the channel navigation UI 16:41:31 <09g​ammafunk> I feel that could get out of hand 16:46:55 <10P​leasingFungus> me on seeing a couple of threads 16:46:56 <10P​leasingFungus> https://cdn.discordapp.com/attachments/747522859361894521/870044675984220300/image0.jpg 16:47:40 <09g​ammafunk> I don't like threads...they're coarse and rough...and they get everywhere! 17:02:38 hello, I have a small bug to report 17:02:49 I died as a FeHW with frozen ramparts up 17:03:04 on revival the frozen ramparts are still present on the level (and don't seem to be going away) 17:03:16 <06a​dvil> nice 17:03:37 <06a​dvil> it looks like archiving that test thread dropped it from the channel list at least 17:03:42 <10P​leasingFungus> very good bug 17:03:55 <10P​leasingFungus> i wonder what happens if you die while channeling mcc 17:04:10 :) 17:04:16 I'm glad it is providing some pleasure 17:04:36 a small update, I just cast ramparts again and now both sets are up 17:06:43 <09g​ammafunk> ho boy 17:06:59 <09g​ammafunk> I probably didn't consider a case like that in the ramparts expiring code 17:07:03 <09g​ammafunk> I can take a look in a bit 17:07:36 <09g​ammafunk> although I hate to ruin the idea of a felid megazigger that managed to use that to completely rampartsify a level 17:11:51 <09g​ammafunk> Ok, with that CUE is added to the tournament configs and we now have all servers except CWZ and CDO available. Rules page should correctly describe all rules we're going with https://crawl.develz.org/tournament/0.27/ . The summary https://crawl.develz.org/tournament/0.27/overview.html shows data taken from 0.27 games since the July 27th (27!). Servers installed 0.27 at different timeframes so won't 17:11:52 be evenly represented. Let me know @ebering if anything looks amiss or likewise to anyone else if you notice anything. Haven't heard back from hong nor Napkin, but dilly said he'd ask hong, so hopefully we'll be good there. And if we don't have CDO itself with 0.27, not too big a deal. 17:13:21 <10P​leasingFungus> excellent 18:00:00 New branch created: pull/2045 (1 commit) 13https://github.com/crawl/crawl/pull/2045 18:00:00 03Elliott Bernstein02 07https://github.com/crawl/crawl/pull/2045 * 0.27-a0-395-gdc5a2530dc: move scroll of noise from bad item to dangerous item 10(5 minutes ago, 1 file, 2+ 2-) 13https://github.com/crawl/crawl/commit/dc5a2530dc4b 18:12:54 <09g​ammafunk> wonder if that's elliott 18:12:57 <09g​ammafunk> ??elliott 18:12:57 <04C​erebot> elliott[1/8]: disregard the elliott 18:14:08 Have you removed spells lately? https://github.com/crawl/crawl/issues/2044 18:23:34 Unstable branch on underhound.eu updated to: 0.28-a0-46-g7b1be97da7 (34) 19:04:00 <05k​ate> useful 20:01:40 <09g​ammafunk> @advil I've gotten two reports that numpad 5 key no longer works for players in webtiles. Not sure how widespread the issue is. Presumably related to your recent commit 20:53:09 do you know what browser / server? 20:54:04 <09g​ammafunk> I'll see if I can find out 20:54:50 <09g​ammafunk> Looks like gosagan was having the same issue on cko, but no word on browser 20:54:54 works ok for me in firefox, chrome / cko 20:55:44 <09g​ammafunk> &rc gosagan 20:55:46 <04C​erebot> https://crawl.kelbi.org/crawl/rcfiles/crawl-0.26/GoSagan.rc 20:56:02 <09g​ammafunk> !lg gosagan x=cv 20:56:04 <04C​erebot> 35. [cv=0.26] GoSagan the Severer (L13 MiBe of Trog), slain by a meliai (a +0 hand axe) on D:15 (hangedman_depths_entry_civil_slaughter) on 2021-07-25 23:56:17, with 35036 points after 15801 turns and 1:02:14. 20:56:51 <09g​ammafunk> I've asked both users about server/browser, so maybe one will respond 20:57:22 I guess if they could it would also be useful to know what keycode the macro entry menu reports for that key, and whether other numkeys are working 21:16:12 i couldn't reproduce this when i opened the savefile again. https://github.com/crawl/crawl/issues/2044