IRC logs for #openttd on OFTC at 2026-01-06
⏴ go to previous day
00:09:32 <_glx_> commit does too many changes to properly track improvements
01:02:11 *** tokai|noir has joined #openttd
01:02:11 *** ChanServ sets mode: +v tokai|noir
01:09:05 *** tokai has quit IRC (Ping timeout: 480 seconds)
01:13:55 *** Flygon has quit IRC (Read error: Connection reset by peer)
01:44:40 *** MinchinWeb[m] has quit IRC (Read error: Connection reset by peer)
01:46:54 *** MinchinWeb[m] has joined #openttd
02:24:51 <ragin_> reldred: I did that ages ago lmao
02:25:17 <ragin_> it's called Auztrees and it's on Bananas
02:41:25 <sittinbythefire> Hmm, maybe I should release a stealth update to ARRS to add the new overlay sprites to help prevent the issue in #15038? It seems like a fair few people have continues to cling on to the ARRS stuff which I suspect may cause more issues like this one.
02:56:58 *** MinchinWeb[m] has quit IRC (Ping timeout: 480 seconds)
03:00:48 <sittinbythefire> Ooh, as a bonus I just noticed the new bridge sprites are missing from the type table in the action 5 docs. Guess I'll add that to the todo list 😛
03:08:27 *** Wormnest has quit IRC (Quit: Leaving)
03:12:00 *** MinchinWeb[m] has joined #openttd
03:39:30 *** emperorjake has joined #openttd
03:39:43 <emperorjake> ragin_: No tropic support though so time for abuse from reldred 😈
03:40:04 <reldred> yeah that's what I meant when I said I wanted to do one
03:40:14 <reldred> I wanna split the trees up for the three tropic biomes properly
03:41:01 <ragin_> well iirc it's in my github if you wanna do it yourself 🙂
03:41:10 <ragin_> wouldn't be too hard tbh
03:41:36 <reldred> nah trees aren't hard
03:41:47 <reldred> it's just sprite replacing
03:47:32 <ragin_> it's why I was able to do it lmao
04:47:23 *** MinchinWeb[m] has quit IRC (Ping timeout: 480 seconds)
04:48:43 *** MinchinWeb[m] has joined #openttd
04:48:46 <DorpsGek> - Update: Translations from eints (by translators)
06:56:19 *** ChanServ sets mode: +v tokai
06:57:18 *** tokai|noir has quit IRC (Ping timeout: 480 seconds)
07:13:09 <_zephyris> sittinbythefire: Sorry, thought I'd done that
07:59:10 <_zephyris> Done now - I'd done the description but not the table!
08:22:38 *** WormnestAndroid has quit IRC (Ping timeout: 480 seconds)
10:15:30 <LordAro> ...why did i just look out of my window?
10:39:49 <_zephyris> I thought scotland was supposed to get snow...
10:53:01 <kuhnovic> The entire train network was shut down here in NL
10:53:17 <Rubidium> kuhnovic: you're so dramatic...
10:55:19 <Rubidium> it's only NS that didn't run, Keolis did run just fine this morning between Amersfoort and Ede
10:59:51 <kuhnovic> "Only NS" is the vast majority of the country. But yeah, if you have access to another rail provided then you got lucky.
11:01:12 <Rubidium> oh, I didn't take that train... just watched it go by :D
11:03:15 <Rubidium> it's just that there's some nuance; as "entire network" vs "only one rail company didn't run" are both technically true
11:14:13 <kuhnovic> Fair enough. I just did my usual say-something-then-think
11:32:51 <peter1138> Sometimes I don't bother with the thinking bit at all.
12:13:43 <xarick> visual studio is becoming broken
12:14:21 <xarick> conflict / merge editor doesn't detect the conflict
12:15:58 <xarick> "click to resolve conflict" - 0 conflicts
12:16:31 <xarick> "Accept merge" - conflict disappears, like... it was magically solved
12:29:02 *** WormnestAndroid has joined #openttd
12:38:14 *** Smedles_ has joined #openttd
12:40:02 *** Smedles has quit IRC (Read error: Connection reset by peer)
12:58:52 *** gelignite has joined #openttd
13:17:57 *** Smedles has joined #openttd
13:20:58 *** Smedles_ has quit IRC (Ping timeout: 480 seconds)
13:31:22 <andythenorth> it was a good time to release also
13:43:47 *** Flygon has quit IRC (Remote host closed the connection)
13:57:52 <pickpacket> Do any of the new age verification laws affect OpenTTD?
13:58:15 *** Wormnest has joined #openttd
13:58:17 <_glx_> we don't store any user related data
13:58:41 <pickpacket> _glx_: is that all the laws are about?
13:58:49 <pickpacket> I don't really know
14:38:28 *** MinchinWeb[m] has quit IRC (Read error: Connection reset by peer)
14:38:42 *** MinchinWeb[m] has joined #openttd
14:57:13 <peter1138> If we ran all the multiplayer servers then definitely, but we don't.
15:00:55 <_glx_> the assembly for this function is so hard to follow
15:01:46 <_glx_> at least I could spot the != Invalid() check
15:02:18 <_glx_> but I could also be wrong
15:02:24 <LordAro> group parent not getting set to Invalid when reordering perhaps?
15:02:48 <LordAro> strange to crash on initial load like that though
15:02:56 <LordAro> perhaps a psuedo-desync?
15:06:14 <xarick> _glx_ is this in a better state? PR?
15:08:55 <_glx_> yes, and with perf test for each commit to see exactly where the gains are
15:09:49 <_glx_> because I'm quite sure some of the changes won't have a big impact
15:11:04 <xarick> gonna need some 5 hours for benchmarks, better get ready
15:15:59 <xarick> hmm do i have to test each change separately, can they be cumulative?
15:16:16 <xarick> commit 2 will include commit 1, commit 3 will include 2+1?
15:20:30 <_glx_> xarick: cumulative is fine, as long as it shows the effect of the new commit
15:22:27 <_glx_> oh the nice effect of removing asserts for stable release ?
15:24:13 <peter1138> (Rect) (t=0, l=2147476352, b=0, r=-2147467872)
15:26:41 <_glx_> so it should have failed in a RebuildXXXKdtree() call before the UpdateGroupChildren() call?
15:28:41 <xarick> something about sign relocating?
15:28:51 <xarick> if it's that, then it's not me
15:29:07 <peter1138> _glx_, no, the kdtree is just a symptom.
15:31:00 <xarick> I'd like to stay and chat, but...
15:31:00 <xarick> that's a lot to bench :(
15:31:07 <peter1138> (const Viewport &) 0x00007fff751968f0: { left = 0 top = 0 width = 1920 height = 1371 virtual_left = 2147483647 virtual_top = 2147483647 virtual_width = 61440 virtual_height = 43872 zoom = Out8x
15:31:22 <peter1138> virtual_left and virtual_top seem unlikely.
15:40:54 <peter1138> So it's a 13.4 savegame initially.
15:57:02 <_glx_> anyway that could mean missing conversion on initial load
16:03:58 <peter1138> _saved_scrollpos_x is INT_MAX if there is no mian window when the game is savaed.
16:04:09 <peter1138> no main window when the gmae is saved.
16:05:22 <peter1138> Ah, okay, this secondary issue is probably just because I'm loading the crash save from connecting to the server.
16:21:21 *** gelignite has quit IRC (Remote host closed the connection)
16:21:30 *** gelignite has joined #openttd
16:54:57 <peter1138> Well, after ignoring the invalid groups, it does load, but I don't know why it's invalid :o
16:56:07 <peter1138> And I don't know what it's meant to look like.
17:29:56 <andythenorth> "pointless news headline to fill news space" is how I should have read it
17:30:16 <andythenorth> or "fuckwits dismayed by lack of understanding of the world"
17:36:58 *** MinchinWeb[m] has quit IRC (Ping timeout: 480 seconds)
17:38:14 <_glx_> oh the good old 9V battery box
17:38:17 *** MinchinWeb[m] has joined #openttd
17:38:46 <andythenorth> "wii wii wii wii" 🎶
17:38:54 <_glx_> and revert the lamps to make it not blinky
17:40:49 <peter1138> So deleting a group will delete all of that group's children.
17:41:41 <andythenorth> can't find a good video the lego space monorail 😛
17:42:08 <andythenorth> 'toy experts disappointed by Lego automated electronic monorail'
17:42:18 <andythenorth> 'children can just push the train'
17:43:33 <peter1138> Hmm, problems with FlatSet?
17:44:13 *** Wormnest has quit IRC (Ping timeout: 480 seconds)
17:45:33 *** WormnestAndroid has quit IRC (Ping timeout: 480 seconds)
17:46:38 *** WormnestAndroid has joined #openttd
17:46:55 <_jgr_> The way the iteration of child groups is done for deletion looks like it would result in invalid iterator issues
17:47:24 <_jgr_> As each deleted group tries erases itself from the parent
17:48:45 <squirejames> andythenorth: PEAK. I had the airport fire engine with the same lights and stuff. Amazing
17:48:46 *** WormnestAndroid has quit IRC (Read error: Connection reset by peer)
17:51:07 <andythenorth> were play experts dismayed?
17:51:41 <andythenorth> in 1986 you just had to sing "wii wii wii 🎶 " in your head
17:51:48 <andythenorth> in the cold north of England
17:51:54 <andythenorth> where 9V batteries hadn't reached
17:52:47 <belajalilija> 9V batteries??? Outside of fire alarms???
17:53:00 <andythenorth> |wii wii wii wii"
17:54:02 <belajalilija> Weird to think back
17:54:15 <belajalilija> How everything used to take batteries
17:54:25 <belajalilija> Can’t even remember the last time i bought some
17:56:56 <peter1138> My last record preamp used a 9V battery.
17:57:05 <peter1138> Though I did buy it from Maplin.
17:57:40 *** WormnestAndroid has joined #openttd
17:57:55 <andythenorth> Bigtrak ruins play
17:58:04 <andythenorth> children can just imagine a space tank
17:58:58 <andythenorth> ok let's talk about something else, I've falled in a hole 😛
18:11:20 <Rubidium> was about to suggest std::move for performance, but that's ofcourse not going to work as the command is ran at least twice. And probably most groups will have an empty flatset anyhow.
18:14:21 <peter1138> Hmm. Well it probably only needs to be called when Execute is set.
18:14:53 <peter1138> Would miss out a lot of test calls though, which might invite trouble later.
18:15:31 <peter1138> Could check if g->children is empty before copying it?
18:16:40 <_jgr_> If it is empty, copying it is free anyway?
18:17:23 <peter1138> Almost. Probably not worth complicating it.
18:18:07 <peter1138> Thanks for pinpointing this.
18:20:00 *** WormnestAndroid has quit IRC (Ping timeout: 480 seconds)
18:22:24 *** WormnestAndroid has joined #openttd
18:24:03 <squirejames> Andy, you need to stop digging into my childhood memories 😄
18:36:49 *** gelignite has joined #openttd
19:23:30 *** WormnestAndroid has quit IRC (Ping timeout: 480 seconds)
19:23:47 *** WormnestAndroid has joined #openttd
19:38:49 *** Wormnest has joined #openttd
20:24:39 *** gelignite is now known as Guest35279
20:24:42 *** gelignite has joined #openttd
20:29:20 <xarick> I expected more from RemoveMapIter
20:31:15 *** Guest35279 has quit IRC (Ping timeout: 480 seconds)
20:48:26 <xarick> so, commit 1 and 3 are bad, let's undo them
20:50:37 <xarick> the compiler works in misterious wats
21:36:29 *** Wolf01 has quit IRC (Quit: Once again the world is quick to bury me.)
21:41:28 <LordAro> we should make it harder to do that
21:49:25 *** MinchinWeb[m] has quit IRC (Read error: Connection reset by peer)
21:49:58 *** MinchinWeb[m] has joined #openttd
22:00:59 <_zephyris> Nicer when transparent
22:02:19 <talltyler> Looks good, but I’ve always wondered how they get the goods in, without a door 🙂
22:02:35 <talltyler> (Blame Simon Foster for that decision, not you 🙂 )
22:04:23 <_zephyris> Indeed, might update 1x too, a couple of strategic pixels would help
22:04:57 <andythenorth> naptime already 😮
22:20:28 <peter1138> That diesel smoke tho' ;)
22:21:10 <_zephyris> Yeah, forgot about the smoke!
22:22:50 <_zephyris> Added a subtle door
22:23:43 <_zephyris> Hard work to draw, but nice
22:26:46 <andythenorth> HST sprite count in GRFS: now 99
22:27:10 <_zephyris> Yeah, but how many at 4x zoom 😉
22:27:20 <andythenorth> Pikka, Timberwolf
22:27:30 <_zephyris> Welll... good point
22:31:26 <_zephyris> Love Timberwolf's art style
22:33:27 <andythenorth> think Pikka's is 2x also
22:34:31 <_zephyris> Which set is that? Not pineapple...
22:37:12 <reldred> Pineapple didn’t get that far past 1950/1960 I wanna say, set never did get finished,
22:38:10 <_zephyris> Was that when he tried to patreon? kickstarter? fund a set?
22:38:21 <reldred> Nah was back in the TTDPatch days
22:38:46 <reldred> Was the North Queensland themed landscape replacement, industry replacement, houses and train set.
22:39:25 <reldred> Fond memories of a family holiday to kangaroo island and I spent the entire time playing TTDPatch and pineapple set. Was great.
22:40:15 <reldred> On an old Thinkpad x31 I wanna say,
22:44:10 <_zephyris> Interested to see how his latest project goes
22:59:27 <xarick> I expected better results by exluding commits 1 and 3
23:00:40 <xarick> but this is windows... very hard to have nothing running in the background
23:05:04 <squirejames> talltyler: curtain sided?
23:16:41 <xarick> I'm too scared of making a PR
23:23:12 <xarick> ~3.07% / ~13.27% / ~6.06% / ~5.58% faster
23:26:00 <xarick> lazy init is the one that's really worth it all though
23:26:38 <xarick> and yet, binary tree would even do better, except for the valuatetests
23:41:26 *** audigex has joined #openttd
23:41:26 <audigex> _zephyris: Yeah hadn't he lost his job or something? So I guess he was hoping to make a bit of income from newGRF as a side hustle
23:41:26 <audigex> Optimistic given the size of the community
continue to next day ⏵
