IRC logs for #openttd on OFTC at 2024-05-04
⏴ go to previous day
00:03:56 <_glx_> peter1138: not related at all, it's ``` /// a (hopefully) large enough character buffer
00:03:56 <_glx_> std::array<char, 64> number_buffer{{}};
00:09:24 <_glx_> `auto buffer_ptr = number_buffer.begin();` then `*buffer_ptr = '-';` if needed, then `buffer_ptr += n_chars;` (with n_chars being the number of digits), then it's filled backward via some `*(--buffer_ptr) = ...`
00:10:11 <_glx_> could be a compiler bug
00:10:14 <peter1138> Yeah, I'm assuming json.hpp is fine and it's compiler being silly.
00:11:52 <peter1138> Trying to find an SD card reader so I can write a fresh 64 bit Pi OS to it, to then try out Linux Arm64 myself.
00:12:15 *** Xaroth92328 has quit IRC (Remote host closed the connection)
00:15:38 *** Xaroth92328 has joined #openttd
00:16:07 <peter1138> Hmm, this is a RISC OS image.
00:16:10 <peter1138> Not sure I really need that 😄
00:23:24 *** Xaroth92328 has quit IRC ()
00:23:46 *** Xaroth92328 has joined #openttd
00:41:15 <peter1138> Oh, of course, there's no longer a default username and password to login remotely.
00:58:21 *** tokai|noir has joined #openttd
00:58:21 *** ChanServ sets mode: +v tokai|noir
01:04:53 *** tokai has quit IRC (Ping timeout: 480 seconds)
01:09:01 *** ufo-piloot has joined #openttd
01:17:16 *** ufo-piloot has quit IRC (Ping timeout: 480 seconds)
01:18:14 *** patrick[m] has quit IRC (Quit: Client limit exceeded: 20000)
02:17:41 *** gnu_jj_ has joined #openttd
02:20:53 *** gnu_jj has quit IRC (Ping timeout: 480 seconds)
02:41:49 *** Wormnest has quit IRC (Quit: Leaving)
02:44:12 <peter1138> silent_tempest: Yeah, ARM v7l. But only when you boot a 32-bit OS. When you boot the 64-bit version it says it's ARM 8 (but doesn't have the descriptive text.)
02:44:26 <peter1138> That's quite... weird 😄
02:56:50 <peter1138> Generating ... baseset metadata file needs to be optimised, hehe
03:01:29 *** D-HUND has quit IRC (Ping timeout: 480 seconds)
03:04:07 *** dihedral has joined #openttd
04:31:40 *** Smedles has joined #openttd
04:41:13 <DorpsGek> - Update: Translations from eints (by translators)
04:57:15 <merni> peter1138: does someone want to approve 12209 now? 😄
05:18:41 *** salut3585 has joined #openttd
05:18:41 <salut3585> can I use lowerCamelCase ?
05:33:18 <andythenorth> peter1138: chocks away
06:35:55 *** fiddeldibu[m] has quit IRC (Quit: Client limit exceeded: 20000)
07:06:46 *** HerzogDeXtEr has joined #openttd
07:37:45 <salut3585> According to the style description, I can't use Enum this way. But I think it's better this way. Do you think it's possible to break the role's style description?
07:44:20 <salut3585> In this case, we may run the function with unexpected values.
07:44:21 <salut3585> if I follow the style description the code will become unclear
07:48:21 <_jgr_> Have you considered using an existing enum such as DiagDirDiff for this?
07:51:08 <_jgr_> It doesn't make sense to use radians when the only sensible rotations are assigned to the tile grid
08:06:34 <jfs> I think the question here is, what are you trying to achieve in the end? why do you need free rotation instead of only the predefined angles?
08:11:25 <salut3585> maybe DiagDirDiff would be good for this. Thank you
08:12:16 *** kstar892[m] has quit IRC (Quit: Client limit exceeded: 20000)
08:21:36 *** calbasi[m]1 has quit IRC (Quit: Client limit exceeded: 20000)
08:28:56 *** gretel[m] has quit IRC (Quit: Client limit exceeded: 20000)
08:34:50 <peter1138> Is it still compiling?
08:37:03 <peter1138> > ssh: connect to host 192.168.12.14 port 22: No route to host
08:38:36 <peter1138> Stopped at linkgraphjob.cpp. A way to go yet, then.
08:47:11 <tateisukannanirase> I've noticed that sometimes the vehicle `cur_speed` can be >0 when the vehicle is in gridlock of vehicles converging in multiple directions. This seems to be reflected in the UI too. Is there an alternative way to tell if a vehicle is moving or not?
09:04:18 <peter1138> Oh, it crashed again, 15+ minutes ago. Sigh.
09:14:34 <peter1138> Can I run swap over NFS? 😄
09:15:17 <peter1138> Actually, memory. Memory over NFS.
09:16:50 <peter1138> [ 38%] Building CXX object CMakeFiles/openttd_lib.dir/src/network/network_command.cpp.o
09:16:50 <peter1138> virtual memory exhausted: Cannot allocate memory
09:38:18 <peter1138> 1GIB RAM not enough 😦
09:39:04 <truebrain> I noticed that before
09:39:13 <truebrain> also, clang and gcc differ in their requirements 😛
09:39:51 <peter1138> Okay, so Raspberry Pi 5 8GB?
09:40:15 <truebrain> I have a few Pis in my house, I can connect one or two a bit later today
09:40:33 <peter1138> Me too, they're all older 1GB or less though.
09:40:56 <truebrain> now mostly first waiting to see if talltyler received an email I send him 😄
09:41:16 <truebrain> and trying to remember how to manage accounts for various of sites ..
09:42:20 <_jgr_> Sounds like you'd be better off cross compiling than trying to do it actually on the device
09:42:45 <peter1138> Okay, with std::deque,. CreateNewNode() took 518µs.
09:42:54 <peter1138> For 1,000,000 calls.
09:43:04 <peter1138> I think this is fine.
09:44:29 <truebrain> owh, the AWS bill this month was 50 dollar less than last month. That is a nice change.
09:44:39 <truebrain> seems 14.0 does have its effect 😛
09:51:54 <andythenorth> I will send you the bill for my MBP later
09:53:30 <truebrain> ghehe, I do have 8GB Pis here, but I "repurposed" their disks, so I don't have a boot medium for them 😄
09:53:40 <andythenorth> can you boot them from the cloud?
09:53:47 <truebrain> I do not have PXE setup in my network, no
09:54:04 <peter1138> What kind of nerd are you?!
09:54:14 <andythenorth> does your TV have storage?
09:54:17 <peter1138> Wait, why don't I have a PXE setup?
09:54:44 <peter1138> (Do Pis support netbooting? Hmm)
09:54:49 <truebrain> I did find a random 8GB SD .. what will be on it?
09:55:24 <truebrain> okay, so a very slow SD .. that is not going to do any good 😛
09:55:35 <peter1138> Deep in the heart of Yapf. RIP.
09:57:22 <peter1138> I hope Xarick is okay. We've not had a woe-the-game-is-ruined moment for a while 🙂
09:58:57 <andythenorth> I'm sure LC will deliver on that at some point
10:00:19 <truebrain> peter1138: you wanted 32bit of 64bit?
10:00:42 <peter1138> 64 bit, that's what the bug report is against.
10:03:05 <peter1138> So £80 for a Pi 5 8GB... and then add on the a power supply, and a case and... yeah, that's probably about the price of a mac M3 now?
10:03:56 <peter1138> Oh but they are kinda awkward to Linux on.
10:09:04 <truebrain> okay, I found an USB stick, and I am now booting my Pi 4B 8GB 😛
10:10:25 <truebrain> I only don't have an adapter for the HDMI, so I am always just hoping the wifi configuration "just works" 😛
10:13:19 <truebrain> see your DM 😛 Go nuts on it 😛
10:13:38 <truebrain> well, stay on the device; I do not give you permission to scan my network 😄
10:14:11 <_jgr_> I'm still running my Pi Model B, with a half GB of RAM
10:14:14 <truebrain> (I can do too what ever you were trying; I just don't know what you were trying 😄 )
10:14:52 <_jgr_> I haven't tried to build or run OpenTTD on it in recent times 😛
10:15:03 <truebrain> building will fail, we now know 😛
10:18:51 <peter1138> Hmm, this appears to be significantly faster than the Pi 3B.
10:19:29 <peter1138> Well, I'll go with noticably. Maybe not significantly.
10:19:52 <truebrain> 4 cores go brrooooeeemmmmm
10:20:04 <peter1138> cmake optimistically says it's 46% of the way through compiling... after finishing the languages 😄
10:20:18 <peter1138> I have 4 cores too, but with 1GB RAM that doesn't help.
10:25:01 *** reldred has quit IRC (Quit: User went offline on Discord a while ago)
10:28:21 <peter1138> > [ 65%] Building CXX object CMakeFiles/openttd_lib.dir/src/network/network_command.cpp.o
10:29:14 <truebrain> up to 14.3% memory on one of the compile-steps already
10:29:20 <truebrain> nom nom nom, memory, nom nom
10:30:30 <peter1138> Hmm, I guess with an NVMe adapter it won't fit in a standard case.
10:33:01 <truebrain> this Pi is passively cooled atm ... it is hot hot hot
10:33:07 <truebrain> so there is some thermal trottling going on at 😛
10:33:17 <Eddi|zuHause> so it's actively hotted? :p
10:33:28 <truebrain> (I used to have them actively cooled, but I was too lazy to connect that 😛 )
10:56:07 <truebrain> right, added some active cooling
10:56:10 <truebrain> should help with performance 😄
10:56:25 <truebrain> also figured out why emails were not arriving; stupid spam filters 😛 😛
10:56:54 <peter1138> Well, it already finished, so it should be cool now 🙂
10:57:18 <peter1138> Now, why do we 1) not have a cli bootstrap or 2) not have a dummy baseset for dedicated builds...
10:57:45 <truebrain> 2) has been a question for 15 years already 😛
10:57:50 <truebrain> "because nobody did that"? 😄
10:57:59 <kuhnovic> TB put his pi in a bath of mineral oil. Active cooling ftw.
10:58:20 <peter1138> It used to matter when we had a vehicle position hash that relied on sprite sizes.
10:59:33 <truebrain> introducing a lot of cooling at once has an interesting effect on the board
11:00:10 <truebrain> as in, I can hear the PCB "changing size" (what is the right english word .. contract?)
11:00:37 <peter1138> Cool, this build does not crash.
11:00:49 <peter1138> Either we fixed it, or the debug report does not have enough information.
11:02:40 <peter1138> OIkay, so mine is a debug build not relwithdebinfo.
11:02:54 <peter1138> (I figured debug would be useful for, you know... debugging)
11:05:30 <peter1138> Is this Pi OS or proper Debian?
11:08:35 <peter1138> I have no clue but the fact it reports errors in json.hpp for them makes me think it's not doing C++20 right 🙂
11:12:54 <truebrain> It is the RaspOS default OS stuff
11:13:03 <truebrain> I can install an older version if that helps
11:31:35 <truebrain> even if that would allow you to reproduce, the solution would still be to upgrade the OS, I guess 😄
11:31:51 <truebrain> anyway, I assume you no longer need access peter1138 ?
11:32:38 <peter1138> I think you are right, if that's the problem it needs an upgrade. So I guess no longer. Thanks!
11:33:14 <peter1138> Although surprising that it compiles and fails, instead of failing to compile.
11:43:01 *** Flygon has quit IRC (Quit: A toaster's basically a soldering iron designed to toast bread)
11:44:56 <tateisukannanirase> Does anybody know how DestinationID in the order framework can be compared to a RoadStop instance?
11:45:31 <peter1138> Vehicles are ordered to stations, not specific road stops.
11:45:54 <peter1138> So DestinationID will be the StationID.
11:45:55 <tateisukannanirase> Ah, that makes sense.
11:46:11 <peter1138> You can compare the station referenced on the tile instead of going via RoadStop.
11:48:58 <tateisukannanirase> Got it, thanks for the quick reply!
11:56:35 <tateisukannanirase> Found `GetStationIndex(tile)` function 👍
12:03:47 <peter1138> `uint64_t flags; ///< stores which blocks on the airport are taken. was 16 bit earlier on, then 32`
12:03:51 <peter1138> Useful comment at the end...
12:07:42 <_glx_> So max size of an airport is 64 tiles
12:08:12 <peter1138> No, it's the number of blocks, which is a logical group, not each tile.
12:22:45 <peter1138> Hmm, no point touching queue.cpp, that'll be removed 🙂
13:14:37 <rau117> tiny little bug-report aka «i'm lazy to go to github»
13:14:37 <rau117> Openttd survey page have no links to openttd.org, the logo button leads to survey.openttd.org
13:16:11 <rau117> tiny little bug-report aka «i'm lazy to go to github»
13:16:11 <rau117> Openttd survey page have no links to openttd.org, the logo button leads to survey.openttd.org
15:04:43 *** Smedles has joined #openttd
15:12:36 *** Wormnest has joined #openttd
15:51:39 <peter1138> Hmm, servicing intervals are a bit wonky.
16:12:48 <peter1138> Mine were 150 minutes.
16:12:52 <peter1138> Which is a bit too long.
16:18:01 <locosage> or too short depends on how you deal with servicing ;)
16:18:21 <peter1138> Servicing is off, so this only affects autoreplace.
16:19:31 <locosage> ah, yeah, I guess it's kinda long then
16:20:13 <locosage> now when I see "minutes" in openttd my first thought is how long is that in real time 😜
16:21:22 <locosage> because I remember they're suposed to be real but aren't
16:25:53 <peter1138> If it says minutes it means minutes.
16:35:48 *** Smedles has joined #openttd
16:41:37 *** Wolf01 is now known as Guest3450
16:47:48 *** Guest3450 has quit IRC (Ping timeout: 480 seconds)
16:55:22 <_glx_> hmm 27ms -> 37tps, I was confusing the units
16:59:00 <locosage> lol, I though I was confusing them
16:59:03 <locosage> whatever, one of those xD
17:31:29 <peter1138> When you leave 90 degree turns on...
17:36:50 <audigex> Technically it’s probably about 170
17:36:56 <audigex> If we’re being pedants
18:00:19 <peter1138> When we are ever not pedantic here?
18:05:03 <jfs> if you were to build that in TT, it would be a 90 deg turn
18:34:24 *** gelignite has joined #openttd
18:36:28 <truebrain> So glad the joke has to be explained 😛
19:49:17 *** wormnest[m] has quit IRC (Quit: Client limit exceeded: 20000)
20:36:28 <ahyangyi> I love how each wagon hits that buffer stop in the other track
21:49:14 *** keikoz has quit IRC (Ping timeout: 480 seconds)
22:02:37 *** Wolf01 has quit IRC (Quit: Once again the world is quick to bury me.)
22:13:30 *** HerzogDeXtEr has quit IRC (Read error: Connection reset by peer)
22:16:28 <Timberwo1f> Mew calendar saves reveal how long you've been playing a particular save. Not sure how I feel about knowing the number of hours.
22:57:17 <truebrain> Haha, that high of a number 😛
23:47:18 *** nielsm has quit IRC (Ping timeout: 480 seconds)
continue to next day ⏵
