IRC logs for #openttd on OFTC at 2025-10-16
β΄ go to previous day
01:40:56 *** Wormnest has quit IRC (Quit: Leaving)
02:17:56 *** gnu_jj_ has joined #openttd
02:21:34 *** gnu_jj has quit IRC (Ping timeout: 480 seconds)
03:00:04 *** herms2 has quit IRC (Quit: bye)
03:09:35 *** Zathras_1 has joined #openttd
03:09:39 *** Zathras_11 has joined #openttd
03:12:56 *** Zathras_4 has quit IRC (Ping timeout: 480 seconds)
03:13:11 *** Zathras has quit IRC (Ping timeout: 480 seconds)
04:38:53 <DorpsGek> - Update: Translations from eints (by translators)
06:16:20 *** Extrems` has joined #openttd
06:17:15 *** Extrems has quit IRC (Remote host closed the connection)
06:17:15 *** Extrems` is now known as Extrems
07:45:35 *** gelignite has joined #openttd
10:31:00 *** Flygon has quit IRC (Remote host closed the connection)
11:31:39 <xarick> copilot suggested me to use std::array instead of std::tuple π I got worse performance
12:26:04 <peter1138> Just stop asking the hallucination device to hallucinate.
12:29:02 *** alfagamma7 has joined #openttd
12:56:05 <xarick> from 43.6 to 38.9 seconds
12:56:26 <xarick> i tried std::initializer_list, std::array and std::tuple
12:58:30 <rito12_51026> How about structure?
13:00:06 *** Wormnest has joined #openttd
13:00:17 <xarick> copilot gave me a weird idea about storing the 4 corner heights as 1 variable via some key...
13:00:43 <xarick> then use a look up table
13:01:22 <xarick> not sure what exactly it means but... gonna try
13:02:01 <xarick> store the difference 0, 1 and 2
13:07:26 <ahyangyi> What's the difference between that and the 5-bit representation in OpenTTD?
13:16:34 <xarick> interesting, according to it, the std::max computation is no longer required in GetTileSlopeGivenHeight
13:18:30 <andythenorth> if you want to use the word soup robot to improve code then...
13:18:30 <andythenorth> * it's useless on anything more complex than a single method
13:18:30 <andythenorth> * if you trust it to do anything other than make suggestions, youll come unstuck
13:20:59 <xarick> oh wow, it kind of helped
13:28:25 <andythenorth> I have watched GPT generate a lot of code
13:28:51 <andythenorth> often the statistics produce a faster, simpler result than whatever I would have done
13:29:15 <andythenorth> but also, often it produces towering piles of bullshit, adding needless structures and abstractions
13:29:40 <andythenorth> it's not very different to a magic 8 ball
13:41:41 *** vikkeri05 has joined #openttd
13:41:41 <vikkeri05> am i the new terry a davis for making a new song file type?
13:42:15 <vikkeri05> it wasnt that impressive
13:43:17 <vikkeri05> i made it cus i wanted to edit how the stuff sounds without braking the whole sound
13:43:34 <vikkeri05> it uses an algorithm to play the waves
13:43:58 <vikkeri05> its kinda hard to explain
13:44:27 <vikkeri05> im sorry if i told this in the wrong place but i just wanted to share my achievement
13:48:45 <xarick> This was copilot suggestion, I actually saw gains
13:48:57 <xarick> but someone more experienced could make it even better
13:57:09 *** gelignite has joined #openttd
13:59:24 <_glx_> 4 imbricated for loops and a bunch of ifs, not sure about the benefit
14:05:44 <xarick> yeah, some garbage slopes in there indeed
14:06:48 <xarick> gonna try make it not store the fake garbage slopes, brb
14:25:40 <xarick> there is no SLOPE_INVALID π
14:29:24 <peter1138> Hmm, I could waste my money on an MNT Reform...
14:40:47 <xarick> i used 255 to represent an invalid slope
15:10:27 <peter1138> Dear world, stop. I need to get off.
15:18:00 <LordAro> i've spent the last 3 hours inside an ipmi web interface, so yeah
15:18:11 <LordAro> it does not have a | in its keymap
15:46:10 <rito12_51026> xarick: Maybe function as constexpr and array as const instead of static?
15:56:46 <peter1138> static std::array<Slope, 81> _slope_lookup_table = []{
15:56:58 <peter1138> Why the absolute fuck?
15:57:20 <peter1138> Your slop-machine is giving you slop.
15:59:04 <xarick> you mean me inputting the 81 entries manually?
16:00:53 <xarick> 3 height diffs: 0, 1 and 2
16:14:22 <LordAro> peter1138: no, "iKVM/HTML5"
16:14:40 <LordAro> java applets haven't worked without significant effort in years
16:15:06 <peter1138> Which is why mine on the home server is useless :)
16:15:07 <LordAro> eventually, i was able to create a (valid) 1.44MB floppy image and upload that to ipmi, and use that to replace the broken certificate
16:15:34 <LordAro> fuck absolutely everything
16:15:44 <LordAro> especially whoever made the supermicro ipmi interface
16:16:32 <peter1138> Any firmware updates for it? Supermicro are generally considers quite good, I thought.
16:16:46 <LordAro> that was one of the first things i did :)
16:17:09 <LordAro> it's pretty solid as far as stability is concerned, but features... lacking.
16:17:34 <peter1138> I remember dealing with something like this before, where I had to run `loadkeys gb` or something, because the KVM assumes a US layout, and | is in a different place./
16:17:38 <dwfreed> LordAro: the effort is less than you think; I have a Sandy or Ivy Bridge Intel mobo with a java applet; it's annoying, but I now have a script for it
16:18:09 <LordAro> perhaps, but why would i use it when i've got the "native" option?
16:18:11 <dwfreed> The worst part is that it reports the jnlp file is longer than it is, so I have to fetch it with curl which doesn't care
16:18:42 <dwfreed> (and in order to fetch it with curl, I have to steal my session cookie from my browser)
16:18:51 <LordAro> the worst part about the HTML interface is i keep pressing C-w to remove a word in the terminal... which closes the tab
16:19:06 <dwfreed> I hate that about all HTML5 "terminal" interfaces
16:19:09 <LordAro> (and the lack of any copy-paste)
16:20:08 <LordAro> my only other option outside of the floppy disk image was to encode the 7kB binary certificate file as hex and maybe use autohotkey or something to paste it in
16:21:00 <LordAro> well, or actually just go across town with a laptop and a usb stick
17:03:15 <andythenorth> Shall we take up carpentry?
17:09:06 <vondpc> and build a modern rail with wood? π
17:11:47 <rito12_51026> carve out OpenTTD binaries in wood like woodpecker
17:13:11 <peter1138> woodpecker is my ci.
17:14:39 <xarick> then the slope is calculated by just Slope r = _slope_height_diff_table[dn][dw][de][ds];
17:15:12 <xarick> "calculated"... not really, just looked up
17:15:24 <LordAro> you're going to need a *really* compelling reason to use it
17:15:37 <LordAro> and no, "saves 5s during river gen on a 4k map" is not even close to enough
17:16:44 <LordAro> (leaving aside the ridiculous unnamed 0xFF Slope constant)
17:26:10 *** tokai has quit IRC (Ping timeout: 480 seconds)
17:27:53 *** ChanServ sets mode: +v tokai
17:28:26 <rito12_51026> Simple question, If I were to remove m6, m7 and m8 from tile structure should I then convert tiles from old saves in afterload.cpp or saveload.cpp?
17:38:34 <xarick> can't extract more performance
17:39:05 <LordAro> rito12_51026: you'd need to do conversions for the MAP* chunks, i imagine
17:39:14 *** tokai has quit IRC (Ping timeout: 480 seconds)
17:39:22 <LordAro> but if it's a 1-1 mapping, maybe it's just a case of updating all the names...
17:45:50 <peter1138> andythenorth, it's not just that. The guy force merged the slop directly to master, bypassing things like CI checks.
17:47:22 <peter1138> Not really relevant, but mildly, he was also refused to change the default texture filtering away from bilinear, which is the reason gzDoom always looks like shit.
17:48:27 <peter1138> Once of the first things changed in the fork :p
17:48:46 <andythenorth> imagine people just committing to master π
17:48:49 <andythenorth> or even in svn π
17:50:31 *** ChanServ sets mode: +v tokai
17:52:06 <_glx_> reminds me when we introduced commit style enforcing
17:53:20 <andythenorth> Anyway open source projects are prone to sort of blowing up, weβve managed not to so far π
17:53:33 <andythenorth> Hoping it continues
17:53:44 *** tokai|noir has joined #openttd
17:53:44 *** ChanServ sets mode: +v tokai|noir
17:54:18 <cu-kai> well the gzdoom drama was a fun read
17:55:42 <peter1138> _glx_, I think it didn't take us long to get used to need approvals.
17:56:01 <talltyler> The article makes it sounds like the AI code was the final straw.
17:56:01 <talltyler> > But the new project will also provide "a more stable development structure with healthy collaboration and less power given to individual 'project leads,'"
17:56:57 <peter1138> It kinda was, but not the only straw.
17:57:02 <_glx_> yeah and I think PR is the best thing (even if we can still miss obvious mistakes when reviewing)
17:58:04 <_glx_> anyway even before the move to github we usually already submitted patches before merging
17:58:20 <peter1138> 2018 was the last svn commit :)
17:58:34 *** tokai has quit IRC (Ping timeout: 480 seconds)
17:58:39 <rito12_51026> LordAro: I think I will store that data temporarily in a sub-tile and make m6, m7, m8 functions refer to sub-tile's m3, m4 and m2 respectively. They will create sub-tile on first call. It will be removed after the save has been loaded.
18:01:40 <peter1138> rito12_51026, I would suggest read the room on how massive changes get implemented. I suspect you are going to end up with a huge patch that nobody is going to want to review. And that's if you're even working on a goal in the way that is desirable.
18:04:47 <rito12_51026> What do you mean by room, could you elaborate?
18:15:33 <kuhnovic> It would suck to spend a ton of time on it if it doesn't get approved. And right now I'm not sure people think this is a good idea. Too many invasive changes for too little gain.
18:15:54 <kuhnovic> (I haven't really looked into it myself yet, I don't have an informed opinion at this point)
18:18:15 <talltyler> I don't even know what you're working on. Getting some idea whether your goal is a desired contribution would be a good place to start. π
18:19:59 *** tokai|noir has quit IRC (Ping timeout: 480 seconds)
18:22:58 *** ChanServ sets mode: +v tokai
18:34:17 <peter1138> Yeah, that's a double-attack of hugeness.
18:35:47 <pickpacket> I've had an Outstanding rating on a station for more than 20 years now and the production of the connected industry doesn't go above 72 units/month. It's really weird. It started at 45, but for the past 15 years it's just flipped between 64 and 72. Smooth economy settings
18:37:48 <peter1138> The flipping is because the production cycles don't align with the calendar nor economy months. It's not actually change the rate-per-production cycle.
18:40:10 <pickpacket> yeah, it just seems so unlikely that production would go *down* as often as it goes *up* considering the rating
18:40:21 <pickpacket> but it just popped up to 90 now!
18:44:56 <peter1138> It's not going down, nor up. It's staying the same.
18:46:29 <peter1138> The production is, in that case, "8". Because the cycles are not aligned, some months you have 8 production cycles, and other months you have 9 production cycles.
18:51:37 <rito12_51026> rito12_51026: Just realised that I have been doing it wrong and now I have to revert my changes
18:59:21 *** tokai has quit IRC (Ping timeout: 480 seconds)
19:03:58 <peter1138> 13:09:53 <peter1138> Why did I have a branch that extends railtypes to 32...
19:04:02 <peter1138> Famous last words...
19:06:02 <xarick> the Encoded Slope Key method is still the fastest
19:06:07 <peter1138> 22:06:05 <peter1138> What about if there was 16 rail types on the map, but more behind the scenes?
19:06:10 <peter1138> 22:06:18 <peter1138> Basically an indirection layer
19:06:16 <peter1138> Indirection layers eh?
19:06:35 *** ChanServ sets mode: +v tokai
19:07:00 <peter1138> Fortunately nobody said anything about #14648, so I can deleted it again.
19:09:22 <rito12_51026> kuhnovic: I'm not sure whether I mind that. I get huge dose of happiness just from writing the code, I prefer the process than the result.
19:15:53 <andythenorth> peter1138 I think 32 railtypes would be great!
19:15:58 <andythenorth> why don't you PR it
19:17:36 * andythenorth me lights a memorial candle for 14648
19:19:04 <locosage> 64k railtypes, one railtype per object :p
19:23:20 <andythenorth> then we can have accurate milepost sprites
19:23:33 <andythenorth> one railtype per milepost
19:24:40 <rito12_51026> andythenorth: Wouldn't randombits be better for that kind of thing?
19:26:13 <xarick> 10 days till I'm locked out of github π
19:26:55 <andythenorth> no, to create my realistic map of the UK rail network on a 4kx4k map, I need 65000 milepost sprites
19:27:13 <andythenorth> it just might have accessibility issues?
19:28:03 <andythenorth> you don't have a TOTP app?
19:28:22 <andythenorth> do you have your own device?
19:28:27 <andythenorth> or is it shared?
19:34:06 <xarick> I don't know what I need as a standalone device
19:35:03 *** orudge` has joined #openttd
19:35:28 *** tokai has quit IRC (reticulum.oftc.net helix.oftc.net)
19:35:28 *** reldred has quit IRC (reticulum.oftc.net helix.oftc.net)
19:35:28 *** belajalilija has quit IRC (reticulum.oftc.net helix.oftc.net)
19:35:28 *** ahyangyi has quit IRC (reticulum.oftc.net helix.oftc.net)
19:35:28 *** kuhnovic has quit IRC (reticulum.oftc.net helix.oftc.net)
19:35:28 *** forgenet has quit IRC (reticulum.oftc.net helix.oftc.net)
19:35:28 *** yiffgirl has quit IRC (reticulum.oftc.net helix.oftc.net)
19:35:28 *** jfkuayue has quit IRC (reticulum.oftc.net helix.oftc.net)
19:35:28 *** moll has quit IRC (reticulum.oftc.net helix.oftc.net)
19:35:28 *** SpComb has quit IRC (reticulum.oftc.net helix.oftc.net)
19:35:28 *** orudge has quit IRC (reticulum.oftc.net helix.oftc.net)
19:36:40 *** reldred has joined #openttd
19:36:40 *** ahyangyi has joined #openttd
19:36:40 *** kuhnovic has joined #openttd
19:36:40 *** helix.oftc.net sets mode: +vov tokai orudge orudge
19:36:56 *** orudge has quit IRC (Ping timeout: 480 seconds)
19:36:58 *** moll has quit IRC (Remote host closed the connection)
19:37:21 *** tokai has quit IRC (Ping timeout: 480 seconds)
19:37:21 *** reldred has quit IRC (Ping timeout: 480 seconds)
19:37:21 *** ahyangyi has quit IRC (Ping timeout: 480 seconds)
19:37:21 *** kuhnovic has quit IRC (Ping timeout: 480 seconds)
19:37:23 *** tokai|noir has joined #openttd
19:37:23 *** ChanServ sets mode: +v tokai|noir
19:37:38 *** ahyangyi has joined #openttd
19:37:39 *** kuhnovic has joined #openttd
19:38:03 *** reldred has joined #openttd
19:38:09 *** jfkuayue has joined #openttd
19:38:09 *** forgenet has joined #openttd
19:38:10 *** yiffgirl has joined #openttd
19:38:24 <xarick> every authentication app i look at have a phone background in mind
19:38:45 *** belajalilija has joined #openttd
19:39:22 <LordAro> maybe it'll force you to start reading
19:48:40 <xarick> all that stuff to me reads "you need a phone"
19:50:26 <peter1138> > You can configure two-factor authentication (2FA) using a TOTP app on mobile or desktop
19:51:20 <talltyler> Step by step instructions for finding a desktop TOTP app, including what to Google π
19:51:34 <talltyler> I use 1Password, but it's not free
19:54:30 <_glx_> a simple google search points me to free apps in microsoft store
19:56:46 <xarick> Microsoft Authenticator is not available for PC or Mac as authenticator apps are typically designed for smartphones for security reasons.
19:58:05 <dwfreed> There are other options
19:58:46 <xarick> 1password wants money and my passwords... srlsly
19:59:22 <dwfreed> a password manager generally wants to store passwords
19:59:31 <dwfreed> do you have a brain, or is it just air up there?
20:01:54 <andythenorth> I use 1password for moneuy
20:02:48 <xarick> 2fast - i dunno what im downloading, but k...
20:03:15 *** tabytac has joined #openttd
20:03:15 <tabytac> I've heard of 2fast before
20:03:20 <andythenorth> "it's only creds, just download any random one"
20:03:36 <andythenorth> "the best ones are from north korea"
20:12:53 <peter1138> keepassxc is the one true password manager.
20:13:08 <peter1138> Everything else seems to pivot to PasswordsAsAService.
20:14:06 <dwfreed> one of these days I'll get around to deploying vaultwarden
20:14:24 <peter1138> Apparently 1Password is "worth" $6.8 BILLION.
20:14:33 <peter1138> So yeah... that's a load of a bullshit.
20:14:34 <_glx_> I use keepass for passwords
20:14:46 <dwfreed> I use a yubikey for TOTP
20:15:06 <_glx_> and google authenticator for totp
20:15:27 <peter1138> But I don't expect them to know how a password manager works.
20:18:21 *** Flygon has quit IRC (Remote host closed the connection)
20:18:55 <_glx_> 2fast seems to be a nice app
20:22:12 <andythenorth> seems like it's closer to naptime than lunch?
20:31:07 *** michi_cc has joined #openttd
20:31:07 <michi_cc> rito12_51026: IMHO, my branch probably needs about 20-30 more commits before changes to the `Tile` structure make any sense.
20:34:32 <rito12_51026> That implicit constructor coused crashing in the mainmenu
20:39:03 <rito12_51026> Removing it will help with rebasing
20:40:22 <xarick> 2fa enabled with KeePassXC
20:40:58 <xarick> why is this even a thing
20:41:39 <_glx_> now you have a bonus, you can generate a different password for every account and store them in keepassxc
20:42:45 <dwfreed> because having unique passwords for different accounts prevents compromise at one service from affecting your accounts at other services
20:43:05 <_glx_> but 2FA is important because many users use a single password for multiple accounts
20:43:54 <_glx_> and many services are unable to properly secure their own data
20:44:23 <dwfreed> I just got an email from Have I Been Pwned yesterday that my info was in yet another compromise
20:51:27 <xarick> now it asks for the password for the kbdx file
20:51:50 <cu-kai> lol, you mean to say you actually use a microsoft account?
20:51:54 <dwfreed> you can presumably tell keepassxc not to run at startup if you don't want
20:52:15 <xarick> i use a microsoft account
20:52:23 <cu-kai> i am sorry to hear that
20:52:26 <cu-kai> local accounts are best
20:53:02 <xarick> i use local account, i mean...
20:53:28 <xarick> it just boots straight to desktop, no PIN, no password asked
20:53:43 <xarick> it's the KeePassXC that wants a pw for the file
20:59:47 *** Wolf01 has quit IRC (Quit: Once again the world is quick to bury me.)
21:04:24 <andythenorth> of the first time I saw a password manager
21:04:38 <andythenorth> seemed like a mad idea
21:06:30 <xarick> can I uninstall KeePassXC now
21:06:42 <xarick> or is github gonna annoy me somehow
21:06:43 <andythenorth> no, you'll need it to TOTP github from time to time
21:07:18 <andythenorth> it sucks, until you find you have employees or family members sticking passwords on post it notes
21:07:32 <andythenorth> or the cover of their notebook
21:07:35 <xarick> I don't have family members
21:07:56 <xarick> I'm basically alone now
21:08:20 <mmtunligit> i loved moving to a password manager, i only have to remeber two passwords now, my computer password and my manager password
21:08:45 <mmtunligit> everything else i can just copy paste, and stuff on my browser gets one click autofilled if im logged in to the manager extension
21:08:52 <xarick> I tried removing the password but it's forcing me to add something...
21:09:00 *** keikoz has quit IRC (Ping timeout: 480 seconds)
21:09:23 <mmtunligit> 2FA was just a bonus that i basically never use because im lazy
21:09:56 <xarick> i dunno what it wants because somebody decided it was okay to run the program in a badly translated portuguese
21:10:04 <xarick> and the translation sucks
21:10:35 <mmtunligit> is your system language portugese? it might be defaulting to that
21:10:36 <peter1138> Tools -> Settings -> Basic Settings -> Automatically launch KeePassXC at system startup
21:12:42 <peter1138> Wild how lax someone can be about basic security.
21:13:01 <xarick> database blocked because i restarted windows
21:13:17 <reldred> Yeah you have to re-open it each time
21:13:31 <reldred> Wouldnβt be much of a password vault otherwise lmao
21:14:04 <xarick> doesn't even remember my password
21:15:27 <reldred> Why would I want the password vault, where I keep my Digital Crown Jewels, to automatically open itself for *anyone*???
21:15:32 <andythenorth> I remember when I was like....why would I have a password on my mac?
21:15:46 <andythenorth> I was really grumpy about having to get digitally paranoid
21:16:08 <andythenorth> then I fell in some tinfoil hole where I didn't trust the password manager either
21:16:22 <andythenorth> so I was authenticating every time I wanted a password
21:16:29 <andythenorth> and using copy-paste, not autofill
21:16:33 <andythenorth> guess how well that goes?
21:16:40 <mmtunligit> i mean it does sound fishy, so now i have a single point of failure for everything?
21:16:51 <xarick> well, I hope it's only github that forces this crap on me... I'm not amused
21:16:51 <mmtunligit> but i trust the security professionals if they say its good
21:17:10 <andythenorth> and if you have a paranoid developer friend, who explains that quite possibly the password app is malware
21:17:19 <andythenorth> because it's not got verifiable chain of integrity
21:17:20 <mmtunligit> and now have memorized a long random alphanumeric/special character string
21:17:24 <andythenorth> and that your laptop might be spying on you
21:17:39 <andythenorth> xarick: I have bad news about other services
21:17:46 <reldred> Ah yes the spooky evil intel management engine
21:18:10 <andythenorth> I wasted quite a lot of time in tinfoil land
21:18:24 <andythenorth> meanwhile I pasted my password to public channels about 10 times
21:18:37 <andythenorth> because having passwords on the clipboard is a Great Idea
21:19:12 <andythenorth> can I go to sleep yet?
21:19:24 <andythenorth> I think I have an ill, my throat hurts
21:19:40 <reldred> keepass and keepassxc will handle automatically clearing clipboard just so long as you donβt go and immediately paste from your password manager straight into your shell session with irssi running
21:19:42 <mmtunligit> andy youre allowed to sleep whenever you please, were just people on the internet, we have no say over you
21:19:46 <mmtunligit> hell im not even real
21:20:25 <andythenorth> I do a lot of deciding, so it's nice to delegate it to the crowd sometimes
21:20:31 <reldred> I use keepassxc and keepassium on ibone, and then regular keepass at work because application whitelisting blah blah blah
21:20:39 <andythenorth> everyday I decide to clean my teeth
21:20:39 <mmtunligit> andythenorth: felt
21:20:41 <andythenorth> that sort of thing
21:20:47 <mmtunligit> everyday i wake up
21:20:53 <mmtunligit> *slams fist on table*
21:27:28 <xarick> well database got corrupted...
21:27:53 <peter1138> It says that if you use the wrong password.
21:30:03 <xarick> ah, i managed to setup that key-file thingy
21:30:10 <peter1138> The password is a secret key to unlock the database. There is no difference between a invalid key and opening a random or corrupted file.
21:30:40 <xarick> at least the key-file is remembered
21:31:08 <xarick> with a key-file it is allowing me to put no password
21:31:17 <andythenorth> xarick: can co-pilot explain a password manager to you?
21:31:23 <xarick> much better, i just need to click unblock
21:31:29 *** keoz has quit IRC (Ping timeout: 480 seconds)
21:31:29 <xarick> still annoying but better
21:32:06 <xarick> it's the opposite of fast for sure
21:33:22 <mmtunligit> people with silly opinions on the internet will be here for you when you wake up
21:33:26 <peter1138> A password manager is to keep your passwords safe. Having it permanently unlocked is the opposite of that.
21:35:15 <andythenorth> discord should auto-post relevant xkcds
21:35:25 <andythenorth> there are likely 2 right now
21:35:27 <xarick> i still use a .txt file with passwords in it
21:35:52 <mmtunligit> xarick: ah so a post it on your computer
21:36:07 <xarick> this keypassxc was also added there
21:36:22 <locosage> time to sleep I guess xD
21:36:25 <xarick> thse programs are just annoying
21:37:18 <andythenorth> ^ that was how i got out of the tinfoil "can't prove the USB controller isn't spying on your passwords in RAM" tinfoil shit
21:37:20 <xarick> tomorrow someone will decide a 3fa is better
21:37:28 <andythenorth> again, I have bad news for you
21:37:29 <xarick> so to hell with these programs
21:37:42 <andythenorth> some services I use are 3fa
21:38:24 <peter1138> I recommend just giving up.
21:38:35 <andythenorth> auth, secondary auth via TOTP, then another OTP in another channel for approving specific actions
21:38:54 <jfkuayue> andythenorth: The specific user βabc_rus123β is very something
21:39:13 <andythenorth> bad andythenorth
21:39:23 <andythenorth> they might be quite young and not-native-english-speaker
21:39:30 <andythenorth> but those two excuses are often made
21:39:41 <peter1138> That other person also spouted nonsense about licenses.
21:41:53 <andythenorth> was it more railtypes though?
21:42:00 <andythenorth> I voted, does that get it done?
21:42:28 <andythenorth> one man, one vote
21:42:40 <andythenorth> hmm do railtypes have badges?
21:42:46 <andythenorth> I think there's a PR
21:57:03 <peter1138> Do railtypes have badges? Yes.
22:20:12 <xarick> GenerateWorld function
22:21:02 <xarick> 38.4 was my best result and i can't replicate it anymore
22:25:35 <xarick> if anyone wants to test that...
22:33:46 <locosage> ask copilot to hallucinate some sse4 code π€
22:37:33 <peter1138> top - 23:35:30 up 6 days, 2:49, 4 users, load average: 77.89, 112.17, 57.46
22:37:37 <peter1138> That could've gone better.
22:37:40 <peter1138> 1217970 petern 20 0 265.5g 25.3g 306328 S 0.7 40.3 219:10.42 Microsoft.CodeA
22:37:48 <peter1138> Of course it's Microsoft software... even on Linux.
22:40:40 <peter1138> loadavg went up to 156 :o
continue to next day β΅