01:14:56 -!- amalloy_ is now known as amalloy 01:38:04 Unstable branch on crawl.develz.org updated to: 0.24-a0-582-g4d4af81 (34) 02:10:04 Windows builds of master branch on crawl.develz.org updated to: 0.24-a0-582-g4d4af81 02:58:32 -!- amalloy is now known as amalloy_ 02:59:22 Monster database of master branch on crawl.develz.org updated to: 0.24-a0-443-g80245de 03:11:43 Unstable branch on crawl.beRotato.org updated to: 0.24-a0-582-g4d4af81 (34) 03:30:43 Fork (bcrawl) on crawl.kelbi.org updated to: 0.23-a0-2313-g5eb34e082d 03:57:26 -!- tehgeek is now known as Psymania 05:20:29 -!- TAS-2012v is now known as TAS_2012v 11:23:03 Stable (0.23) branch on underhound.eu updated to: 0.23.1-89-g277498a001 11:27:17 alexjurkiewicz: thank you, i have the graphs on github somewhere 11:50:27 I wonder if it would make sense to have a mailing list for server operators to join, + devteam members who interact with them most? 11:50:53 not sure how much use it would get, since the people most likely to say yes to that already are here regularly 11:52:24 but I wonder if that format might work better for some kinds of server-related communication 11:53:29 (to be clear: which isn't to say at all that I don't think that communication should happen here) 11:54:41 or even clearer: I was thinking about this b/c I was updated some crawl-private mailing list stuff, *not* because the last message in the channel was server-related, that's just a coincidence :-) 11:55:25 it would be v easy for me to host such a mailing list if there is interest 11:56:02 also: any devteam member who was on the old one should have gotten a subscribe message for crawl-private on a new domain, let me know if you didn't 12:06:52 I got mine 12:07:05 ok, thanks 12:10:20 same here 12:11:03 does anyone know if the you.redraw_foo fields serve any purpose other than optimization? 12:11:26 I do not 12:11:34 I was looking at what it'd take to finish refactoring the stats UI code, and I'm considering ripping those conditionals out 12:11:51 though that's the kind of thing where my default assumption would be that removing them would break some corner case I won't discover until the commit 12:12:29 was mostly wondering if they minimize terminal redraw or something 12:13:37 hrm, it's possible I suppose, although the chances of forgetting to set the appropriate flag seem higher to me 12:15:08 yeah, I guess these could be aimed at reducing flicker? 12:15:15 not sure how real a concern that is in 2019 12:16:10 unnecessary redraws still increase server traffic 12:16:27 but the question is whether it's a significant difference 12:16:34 I've played on 3G on trains with noticeable screen redraw although obviously have not performed a comparison with full screen every time 12:16:49 afaik ncurses does some diffing / buffering; a quick look at the w32 code suggests it does the same 12:17:05 I notice from ipbt replay that Crawl seems to spit out a lot of frames each move 12:17:36 (Compared to, say, NetHack which is more or less "here is the next frame" if you don't have messages to page through.) 12:17:52 it also seems entirely plausible to me that there could be many cases that someone has ended up forcing a full redraw, undoing any effect of this optimization 12:18:08 aidanh, yes, ncurses diffs the screen contents 12:18:42 on os x there is a neat utility that allows you to simulate terrible network traffic, using something like that might help in seeing what's going on 12:19:19 "Network Link Conditioner", installed via xcode 12:19:29 something probably exists for other OSs too? 12:20:21 %git 0a1fc62c7c 12:20:21 07advil02 * 0.23-a0-690-g0a1fc62: Prevent recursion for `viewwindow` 10(8 months ago, 1 file, 90+ 69-) 13https://github.com/crawl/crawl/commit/0a1fc62c7c50 12:21:09 I think that commit is a sign that redraws aren't doing what we think they are at least some of the time 12:25:58 yeah, the current code is nightmarishly complicated 12:26:10 I still haven't figured out how the whole main loop works 12:26:47 i'd suspect always doing a full redraw is probably safer, in that it'd prevent the stats sidebar from ever getting out of sync 12:48:03 there's a lot o forced redraws just so ttyrecs are saneish 12:48:34 and yes, theres at least one linux kernel module that simulates various kinds of network issues 12:48:46 not bundled, I think 13:13:02 Unstable branch on crawl.akrasiac.org updated to: 0.24-a0-582-g4d4af81 (34) 13:18:32 Dispersal trap ends tornado 13https://crawl.develz.org/mantis/view.php?id=12051 by Pekkekke 13:18:52 -!- amalloy_ is now known as amalloy 13:22:52 pekkekke tryng to blinknado 13:23:52 Monster "Blink Other" Ends Tornado 13https://crawl.develz.org/mantis/view.php?id=12052 by Pekkekke 13:40:34 I think those aren't bugs, exactly 13:41:14 they just didn't test all the cases - if you blink out of the winds (no matter how), tornado ends. For some reason this isn't messaged 13:43:12 we should probably go with "always end tornado on tloc" 13:45:58 that seems fine 13:46:02 how about hop? 13:47:59 and I guess wjc stuff 14:03:50 the simulationist in me says "those are different, they're physical motion" 14:04:02 probably fine for wjc since those can all be tripped w/movement keys 14:05:44 simulationists in this channel?!! 14:08:57 right now it looks like the same principles govern each, just depends on distance from previous location (where WJC abilities can't move you far enough to end it) 14:11:08 huh, seems weird to me that blinking out of a tornado stops it, but if you just walk it follows you 14:11:15 it depends on distance 17:54:05 advil: i think i already mentioned i'm glad to see your Morgues advice in the biweekly thread. for the next one, could you also suggest that they make a pastebin/gist of morgue files for ongoing games? 19:59:13 Malfegor (L27 DsHu) ASSERT(range >= 0) in 'beam.cc' at line 641 failed. (beam '', source 'player', item 'none'; has range -1) (Zig:18) 20:10:25 wow, that is an upgrade from quite an old version 20:12:21 hard to be sure without gdb, but i suspect the spell they tried to caste was Haste 20:15:04 yeah, attempting to cast haste crashes. we should probably put a minor version in for it so that in save-compat we can unlearn the spell 20:23:06 hah, and it doesn't even actually have a range of -1. it's default-initialized to -2 21:08:37 -!- amalloy is now known as amalloy_ 23:43:45 -!- amalloy_ is now known as amalloy