IRC logs for #openttd on OFTC at 2023-07-29
⏴ go to previous day
00:47:47 *** HerzogDeXtEr has quit IRC (Read error: Connection reset by peer)
01:07:45 *** Wormnest has joined #openttd
01:07:45 *** Wormnest_ has quit IRC (Read error: Connection reset by peer)
01:54:03 *** Wormnest has quit IRC (Quit: Leaving)
02:13:21 *** Zathras has joined #openttd
02:16:45 *** debdog has quit IRC (Ping timeout: 480 seconds)
02:16:50 *** D-HUND has quit IRC (Ping timeout: 480 seconds)
04:02:25 *** woobilicious has joined #openttd
04:02:26 <woobilicious> `ViewportAddString` seems to be used with less parameters than needed in some places (looking at JGR's branch), and doesn't have any defaults set, isn't that undefined behaviour for C++?
04:13:46 *** Zathras is now known as debdog
06:15:38 <Eddi|zuHause> woobilicious: really? that would sound more like a compiler error than undefined behaviour. probably you're not reading it right.
06:33:21 <woobilicious> Eddi|zuHause: it's pretty hard to mistake passing `t->cache.population` and a colour as the last parameter, counting them would confirm colour is ommited in the first case, there's no default colour set, I think colour is enforced somewhere else in the cases where it's not passed, so this minor issue is probably ignored. but IDK, I've not done much C++ before...just C, so if you could point me to
06:33:21 <woobilicious> where my thinking is wrong I'm open to it, I'm just confused tbh lol
06:34:31 <woobilicious> right now I'm trying to figure out why my new setting `gui.city_in_label`, only half works...console, and openttd.cfg shows it's set to 1, but `printf` is showing it's set to 0.
06:34:38 <Eddi|zuHause> i'm pretty sure colour is "inherited" by some method
06:37:37 <woobilicious> Yeah it would be pretty obvious if town labels had corrupted colours, I'm just supprised it's not a compiler error.
06:50:05 <pickpacket> I’m considering marking stations that are boosted by a statue in town in the station listing somehow. What would be a good symbol for a statue?
06:51:08 <kamnet> Animated rotating OpenTTD logo? 🙂
06:59:47 *** Flygon has quit IRC (Read error: Connection reset by peer)
07:01:54 <woobilicious> Could someone help me understand why this `gui.city_in_label` shows `1`, but `_settings_client.gui.city_in_label` always shows `0`?
07:16:14 *** HerzogDeXtEr has joined #openttd
07:20:10 <andythenorth> pickpacket: 🍭 lollipop for lol gameplay mechanic
07:37:09 <truebrain> We are introducing a new charge for public IPv4 addresses. Effective February 1, 2024 there will be a charge of $0.005 per IP per hour for all public IPv4 addresses, whether attached to a service or not
07:37:30 <truebrain> AWS will get slightly more expensive in next year
07:41:12 <truebrain> Worst part is, that AWS themselves still have not all endpoints available via IPv6 ... like the autoscaling endpoint. Maybe they should fix that first, before "nudging people to use IPv6"
08:03:03 <woobilicious> IPv6 is such a clusterf too, I remember running it in like 2008 on a D-link asdl model I hacked and put openwrt on, but even my more modern router has bugs, due hardware vendor not providing drivers for a version of linux that isn't 10 years old.
08:04:47 <pickpacket> andythenorth: what mechanic? That statue boosts rating? 🙂
08:15:12 *** _zephyris has joined #openttd
08:15:12 <_zephyris> pickpacket: I suggest a generic representation, an icon representing a boost or upgrade: ⏫ Presumably stations can be boosted by other mechanics or by gamescripts, so having a generic boost icon would probably make sense?
08:19:41 <Eddi|zuHause> "boost" like in advertising campaign?
08:26:35 <andythenorth> pickpacket: Yes. It’s one of the sillier game mechanics 🙂
08:47:49 <alfagamma_0007> kamnet: My point exactly!
08:48:15 <alfagamma_0007> At least another one of Sawyer's game at least has weather conditions
08:52:21 <_jgr_> Air resistance is really there because of the v^2 behaviour. You wouldn't get nice acceleration curves at high speed otherwise.
08:57:58 <alfagamma_0007> Also atmospheric conditions are a nightmare to code
09:08:23 <woobilicious> anyone got 5 mins to go over my code and telling me why my new setting doesn't work?
09:48:37 <woobilicious> nevermind figured it out.
10:04:06 <truebrain> Lol, even nginx consumes a nonzero amount of CPU to run the TURN servers ... maybe thr CPU usage of the Python process is not all that unreasonable 😄
10:04:22 <truebrain> It is currently running around 250 requests per second .. lol
10:42:35 <pickpacket> andythenorth: yes, and I love it 😄
10:43:34 <pickpacket> _zephyris: that’s not a bad idea. I wonder what type of symbol would work. Does the UI have unicode support?
10:44:58 <peter1138> woobilicious: That your old shit kit was shit is nothing to do with the state of IPv6
10:45:38 <woobilicious> The state of the "roll out".
10:49:13 <woobilicious> peter1138: most wifi routers are "shit kits", buggy property drivers that only work on old linux kernels and slow overpriced hardware that can't accelerate IPv6 at all.
10:55:28 <peter1138> People will keep buying the cheapest (and not so cheapest but well marketted) shit still...
10:56:29 <peter1138> I don't need IPv6 nor IPv4 acceleration on my WiFi access point as it's just bridged layer 2.
10:59:23 <woobilicious> I should probably just put openwrt on it, I think they have it working correctly. and I don't need accelation either since I'm not running gigabit.
11:03:20 <peter1138> Get a Mikrotik, and never look back.
11:03:30 <woobilicious> It is a mikrotik lol
11:03:52 <peter1138> Okay, then OpenWRT isn't going to make it better.
11:05:13 <woobilicious> have you tried using fq_codel?
11:05:46 <woobilicious> RouterOS 6 doesn't have it, and RouterOS 7 has IPv6 bugs.
11:10:05 <peter1138> No, I don't need any queues... so I can keep Fasttrack active.
11:12:45 <woobilicious> fq_codel should be manditory for for any router,
11:26:20 <peter1138> Okay do I go out and about hunting records (vinyl, not cycling) or just sat here all afternoon binging Good Omens 2?
11:26:39 <truebrain> the choices in life!
12:40:39 *** NGC3982 has quit IRC (Quit: bbl byta bord)
13:17:47 <andythenorth> peter1138: There’s a 2?
13:18:04 <andythenorth> I thought 1 was surprisingly good
13:18:18 <peter1138> Just came out yesterday.
13:34:23 *** belajalilija has joined #openttd
13:34:23 <belajalilija> peter1138: Congratulations!
13:42:56 <truebrain> right, doing some data collection, as .. I am missing that. When a user open the online content window in-game, he gets ~350KB of data 😛
13:43:43 <truebrain> also, what I kinda forgot, for every download it checks on the server if there is an update .. which causes another blob of data 🙂
13:55:48 <peter1138> 350KB is nothing for the MacBook Pro.
14:09:21 <truebrain> Not sure yet these are all useful metrics to monitor, but it seems like a good start. _glx_ if you are a bit bored, could use an extra pair of eyes on that PR 🙂
14:13:34 <_glx_> I like the removal of a left over debug `print` 😉
14:13:51 <truebrain> yeah ... no clue what that was doing there 😄
14:17:11 <truebrain> can you think of other metrics that might give a bit of insight what is going on inside the bananas-server?
14:17:15 <_glx_> at least the additions look logical
14:20:46 <_glx_> of course it's only for non HTTPS stuff
14:21:01 <truebrain> yeah, the HTTP stuff I can see via Cloudflare; I could add it too, but that is a bit more tricky 🙂
14:44:15 <truebrain> enabling prometheus consumes 2 MHz per service 😛
14:44:25 <truebrain> not sure if you wanted to know, but yeah .. now you do 🙂
14:44:38 <truebrain> (2 MHz just for the polling Prometheus does, every 5 seconds)
14:51:03 <Eddi|zuHause> that's like 0.005% cpu?
14:51:12 <truebrain> depends on your CPU, duh
14:51:50 <Eddi|zuHause> let's assume "average business CPU, latest generation"?
14:52:02 <truebrain> how about not making any assumptions?
14:52:39 <Eddi|zuHause> the kind of business where they throw out the whole system after 2 years and get a new one.
15:03:49 <Eddi|zuHause> also, i meant % in the unix sense, i.e. of a single core.
15:04:56 *** NGC3982 has joined #openttd
15:06:31 <truebrain> profing more and more how pointless it is to represent it in % 🙂
15:10:32 <truebrain> odd, BaNaNaS TCP downloads only go with ~20 mbps for me .. should be 1 gbps tbfh .. something to look into, later 🙂
15:10:36 <Eddi|zuHause> every metric is pointless if used outside the meaningful context
15:11:25 <Eddi|zuHause> like, measuring car speed in fractions of c :)
15:17:14 <truebrain> hmm .. the download statistics is weird and seemly wrong ... what is going on there ..
15:21:30 <truebrain> no, it works fine .. it just didn't record my abase download for some reason .. hmmmm
15:25:42 <truebrain> second time it worked .. odd, owh well
15:31:49 <peter1138> A nice decent CPU, like a 6502, would run at 2MHz, although many systems of the time only ran them at 1MHz.
15:32:17 <truebrain> grrrr, how do I make grafana ignore a null entry when calculating the mean
16:07:58 <truebrain> hmm .. if these stats are actual correct, there are far less TCP-based BaNaNaS downloads than I expected .. I just don't trust these numbers just yet .. 😛
16:09:32 <merni> in what circumstance would a bananas download be TCP based or otherwise?
16:09:42 <truebrain> whether HTTP connection can be made or not
16:15:36 <truebrain> there used to be a lot more traffic going via TCP .. but that might have been fixed with the new infra .. or I am doing something wrong 😄
16:16:14 <truebrain> need to find an easy way to get the bnadwidth used by the service itself .. but that is far more tricky then I would like. Owh well, enough toying with stats for one day 🙂
16:18:22 <belajalilija> That’ll be neat for the people who play with AI
17:14:24 *** Venemo has quit IRC (Quit: No Ping reply in 180 seconds.)
17:29:48 *** Wormnest has joined #openttd
17:54:56 <_glx_> oh of course old workflows
18:00:56 <truebrain> yeah, that is totally broken 😛
18:13:31 <truebrain> w00p, I now have combined TCP and HTTP downloads stats for BaNaNaS \o/
18:13:39 <truebrain> now it needs to run for a while to get some decent values 🙂
18:14:04 <truebrain> yeah, something like that 😛
18:14:13 <truebrain> seems TCP vs HTTP is something like 1:24
18:14:35 <peter1138> Does that just mean most people are using 13+?
18:14:44 <truebrain> no, 0.9+ or something
18:14:55 <peter1138> Oh right, so that was ages ago.
18:15:25 <truebrain> and I just never knew how much HTTP actually failed
18:16:19 <truebrain> tomorrow I can give an actual number 😛
18:37:08 <talltyler> truebrain: Need a Steam news image or are we skipping that again?
18:37:25 <truebrain> If you want me to post anything on Steam, you will need to supply me with an image, yes 🙂
18:37:30 <truebrain> but whether we do or not, that is not up to me 🙂
18:37:55 <talltyler> I will make one for _glx_
18:38:31 <DorpsGek> - Update: Translations from eints (by translators)
18:39:56 <truebrain> but needs more data to become a bit more accurate 🙂
18:40:35 <truebrain> (downloads in bandwidth and requests per minute of the last 30 minutes on both TCP and HTTP on BaNaNaS)
18:47:30 <truebrain> Cloudflare agrees with the HTTP numbers .. so I guess the TCP numbers are also valid
19:22:28 <truebrain> and please either make them all end with a dot, or none
19:22:34 <truebrain> but 3 out of the 5 ending on a dot makes me OCD go BRRRRRR
19:22:53 <truebrain> (seems we normally don't have a dot, but it isn't consistent 😛 )
19:23:49 <talltyler> Hmm, lost the actual changelog? I only see lang files now
19:24:25 <_glx_> I'm playing with reflog to fix
19:29:45 <_glx_> did'nt check all the file
19:30:29 <_glx_> note for later, don't run backport script if you add commits to the branch
19:31:01 <truebrain> quick, peter1138 , also approve!
19:31:41 <peter1138> (Should I look at the changes?)
19:31:59 <peter1138> Okay, cos there's a red cross :p
19:32:10 <truebrain> yeah, I broke shit; won't be fixed 😛
19:45:38 <_zephyris> pickpacket: I think it's complicated, for base features you want a pixel font icon/character anyway
19:47:22 <_glx_> unicode is supported, but required glyphs must be in baseset sprite font
19:47:47 <_glx_> at least for english support
19:48:08 <_glx_> other languages may require setting a font
20:24:17 <_glx_> at least the amount of duplicate bug reports should decrease
20:26:51 <_glx_> hmm I hope the release workflow still works for 13
20:27:16 <truebrain> Only now realise your release will fail
20:27:29 <truebrain> We can't release any 13.0 anymore without some big backporting
20:27:49 <truebrain> Owh, you just mention it yourself
20:28:15 <truebrain> Honestly, skipping 13.4 is far easier .. just prep for 14.0 🙂
20:29:08 <truebrain> Bit if you want 13.4, backport all CI related changes 🙂
20:29:47 <truebrain> The red Cross was rightfully predicting what would happen 😛
20:30:43 <_glx_> yeah cdn will definitely be a problem
20:32:04 <_glx_> we should have released right after the backports 🙂
20:32:37 <_glx_> but as always nobody wants to do the changelog 😉
20:42:09 <_glx_> windows store upload passed
20:56:17 <talltyler> _glx_: I can write a website post if you like
20:56:53 <_glx_> well for now I'm waiting to see what pass/fail in the workflow
20:57:16 <_glx_> I already know cdn upload will fail
20:57:49 <_glx_> if gog/steam are fine I think we can work around cdn fail manually
20:58:06 <talltyler> I will write a post then 🙂
20:58:28 <_glx_> yeah I'm not a writer anyway 😉
21:01:27 <truebrain> _glx_: That is a lot of manual labour .. easier to backport the right PRs and start a new release
21:03:17 <truebrain> Better to cancel the run; cleanup up the mess is also very annoying work
21:04:05 <truebrain> _glx_: It didn't upload. It will be copied to the CDN, which will fail. FYI 🙂
21:05:24 *** peter1139 has quit IRC (Remote host closed the connection)
21:05:42 *** peter1139 has joined #openttd
21:06:17 <_glx_> yeah cancelling before it affects steam or gog is a better choice
21:08:29 <_glx_> ok let's see which PRs to backport
21:13:08 <talltyler> Not much writing to be done for this one but 🤷
21:25:11 *** nielsm has quit IRC (Ping timeout: 480 seconds)
21:54:27 <Eddi|zuHause> weird... i got a mail that my microsoft account is about to be deleted from inactivity... and i'm like "what? i had a microsoft account?"
21:55:09 <Eddi|zuHause> but it looks like a genuine microsoft link?
21:57:04 <dwfreed> I got one a while ago
21:57:24 <dwfreed> if you ever had an xbox live account, it got made into a full microsoft account
21:57:57 <dwfreed> you may have signed up back when live.com was a thing
21:59:48 <kamnet> Or hotmail. I still have my original account from 2000-ish.
22:03:18 <Eddi|zuHause> i don't find any stored passwords
22:09:43 <_glx_> hmm I might be including too many changes
22:09:52 <truebrain> you are, but nothing that worries me honestly
22:10:13 <truebrain> updating the linux build might not be really needed, but .. *shrug* .. the 13-variant is a bit weird and broken on Wayland
22:10:30 <truebrain> so one could argue it is a fix that needs backporting 😄
22:11:13 <_glx_> yeah the generic uses very old stuff
22:11:34 <truebrain> so I think your PR is correct, and will work
22:11:43 <truebrain> but as it goes with backports .. we never know till we try!
22:12:36 <_glx_> at least it's building, means cmake was "happy"
22:13:03 <_glx_> can't see cmake output yet as it is still building
22:21:30 <_glx_> at least with the updated workflows, even if the uploads fail because missing credentials, I should still get all the artifacts
22:24:12 <_glx_> cmake output seems fine
22:34:19 <truebrain> the issue is not getting all the artifacts .. the issue is formatting them correctly to upload them, and trigger the right sub-workflows after that 😛 I did it once, I will not do it again 🙂
22:34:26 <truebrain> much easier to fix the workflow and trigger a new release
22:37:43 <_glx_> of course some stuff fails, but it's the expected failures
22:42:59 <_glx_> and linux generic build seems to work fine (I have sound and music from wsl)
22:44:21 <_glx_> no need to update changelog as it's all CI related stuff
22:48:09 <_glx_> (but I think it will be fine)
22:54:35 <_glx_> oh I know, the vcpkg cache
23:16:53 <_glx_> I had to kill linux cache
23:17:09 <_glx_> but it will come back on next nightly
23:17:51 <truebrain> That is perfectly fine
23:18:04 <truebrain> At least you can do that these days from the web
23:21:48 *** keikoz has quit IRC (Ping timeout: 480 seconds)
23:31:49 <_glx_> yeah it's nice to be able to remove "broken" cache
23:35:14 <_glx_> our cache handling is made so newer workflows can start from older cache when upgrading, but the release was the other way and used the newer cache as base to install older stuff (very rare case)
23:41:15 <truebrain> Tomorrow I will press some buttons for steam. Bit late now for me 🙂
23:41:37 <talltyler> Website should be ready to go
23:41:46 <talltyler> Just needs buttons pressed too
23:42:07 <talltyler> (And changelog link verified)
23:42:38 <truebrain> Workflow also failed, so CDN and website won't update atm
23:42:49 <truebrain> Something to fix tomorrow too
23:43:12 <truebrain> The internal folder is an issue (again)
23:43:20 <truebrain> Stupid Windows Store files 😛
23:50:33 *** tokai|noir has joined #openttd
23:50:33 *** ChanServ sets mode: +v tokai|noir
23:50:56 <_glx_> well that part was not really testable during the migration (releases and nightly have some diffs)
23:57:33 *** tokai has quit IRC (Ping timeout: 480 seconds)
continue to next day ⏵
