IRC logs for #openttd on OFTC at 2023-05-31
⏴ go to previous day
02:42:12 *** debdog has quit IRC (Ping timeout: 480 seconds)
03:24:18 *** D-HUND is now known as debdog
05:40:12 *** keikoz has quit IRC (Ping timeout: 480 seconds)
08:44:46 *** _aD has quit IRC (Quit: leaving)
12:55:40 <petern> Hmm, is it ever likely that objects will support cargo variations (a la stations)?
13:01:40 <TrueBrain> lol .. I was all happy that most of the infra is now IPv6-only, as that saves a bit of money .. but .. I forgot we query servers for the master-server, so we need IPv4 connectivity .. hmm .. can I do that in a clever way ... I think I can. Can I? Possibly ...
13:04:26 <petern> Create a bot net of IPv6/IPv4 connected servers to query IPv4-only servers...
13:05:13 <TrueBrain> I was thinking a bit cheaper .. just run a single instance with pproxy, so I can use SOCKS to use that proxy .. that is kinda already running in the current infra. I just have to be smart where to host that. I think.
13:06:06 <LordAro> just run it on your desktop
13:08:36 <TrueBrain> also a bit sad, it seems I cannot route non-HTTP via Cloudflare Access .. meaning I need to use a public IP address on AWS anyway to route our non-HTTP via
13:08:42 <TrueBrain> more infra .. I hate more infra 😦
13:10:27 <TrueBrain> wish we could modify old clients .. would make things easier 🙂
13:13:37 <Rubidium_> are you saying the server list requests should be done via HTTP calls in the future?
13:14:01 <TrueBrain> if we could modify old clients, that would be worth the effort, yes
13:14:05 <TrueBrain> but given we can't, it doesn't matter
13:14:09 <TrueBrain> as we need to support the old way anyway 🙂
13:14:45 <TrueBrain> in this day and age, it is far easier to scale HTTP traffic than any other traffic, basically. So if we could do it again, things like content-service and server-listing would be fully HTTP-based, if it was up to me, yes 🙂
13:16:32 <TrueBrain> but now, I have to launch a 15 dollar per month service just for the non-HTTP traffic .. and that service isn't even the biggest problem, the bandwidth price they charge kinda is 😛
13:20:16 <LordAro> does it need to be AWS? couldn't just find a cheap VPS somewhere?
13:20:30 <LordAro> must be able to do better than $15/month...
13:21:13 <TrueBrain> you can do everything on cheap VPSes, sure, but then we are back to the maintenance nightmare we came from 🙂
13:21:25 <TrueBrain> phrased differently: when is the last time you updated the current 2 VPSes we have? 🙂
13:21:42 <LordAro> we installed unattended-upgrades, didn't we? :p
13:22:30 <LordAro> (i might have done it more regularly if I knew it was something I was expected to do :p )
13:22:54 <TrueBrain> haha, that is one way to word it 😛
13:23:10 <TrueBrain> anyway, let's hope Cloudflare wants to sponsor us; that might give some possibilities 🙂
13:23:28 <TrueBrain> I did send the request, just didn't get anything back yet
13:25:17 <petern> They'll sponsor you if you have a DDoS botnet behind the scenes as that drives their sales...
13:26:26 <h24816> anybody know if the number of newgrf's active in a save affects performance? ie CPU/RAM usage
13:27:37 <LordAro> not so much the number of newgrfs as what they're doing
13:27:59 <h24816> object and building sets mostly
13:28:45 <h24816> my thinking was that the increased number of different sprites available could possibly reduce performance
13:31:26 <TrueBrain> there really isn't a simple answer; it really depends on the NewGRF.
13:32:23 <LordAro> in very simple terms, the answer is "yes"
13:32:31 <LordAro> but you may not actually notice any difference in real terms
13:32:58 <petern> Or you will if the NewGRF uses tile animations
13:34:55 <TrueBrain> so does the number of NewGRFs matter? Doubtful. Does the type of NewGRF matter? For sure!
13:35:05 <TrueBrain> we can create a game with 100+ NewGRFs that runs quicker than one with a single NewGRF 🙂
13:35:16 <TrueBrain> so I disagreee with you LordAro 🙂
13:35:45 <LordAro> but do you disagreeee?
13:35:55 <TrueBrain> even disagreeeeeeeeee
13:36:04 <TrueBrain> too bad you are not joining the 10th
13:36:25 <LordAro> too expensive at too short notice, alas
13:36:49 <LordAro> especially as i'm travelling the week after, and the week after that
13:36:49 <TrueBrain> pff, short notice ..... how much more notice do you need!
13:36:55 <h24816> i see. thanks guys. yeah i should've clarified, i'm mostly referring to non-animated object sets, like fridaemon's and garryg's for example. i actually limit sprite resolution to 8bpp so i assume 32bpp doesn't make much of a difference. but good to know adding a few static object sets won't impact performance significantly.
13:37:21 <petern> 8bpp / 32bpp isn't sprite resolution.
13:38:33 <TrueBrain> LordAro: can I at least take a picture of you with me and put it on a chair next to us?
13:39:05 <Rubidium_> at which station shall we then pick him up? :D
13:43:11 *** h24816 has quit IRC (Quit: Page closed)
13:56:50 *** HerzogDeXtEr has joined #openttd
14:05:46 <FLHerne> petern: how would objects know which cargo was relevant?
14:06:05 <FLHerne> query cargo of a neighbouring station tile?
15:18:45 *** gelignite has joined #openttd
17:48:16 <TrueBrain> if I would be to guess, that was a cache-hit on the website, making it look like that .. mostly as the information comes from the same source 😄
17:48:58 *** Ahyangyi has quit IRC (Quit: User went offline on Discord a while ago)
17:54:01 <TrueBrain> close-as-not-a-bug? 😄
18:01:42 <Eddi|zuHause> was that diplomatic enough?
18:13:26 <frosch> what percentage of users know that the "one industry per type per town" thingie is a setting? and that gs can change settings
18:25:09 <Eddi|zuHause> my standard answer for settings is: 1% :)
18:27:04 <TrueBrain> frosch: "soon" we can answer how many change it 😛
18:36:08 <frosch> ah, a scheme->wasm compiler. scheme may just be the on the right wtf level for people who learned newgrf before 🙂
18:38:19 <TrueBrain> so for you, you mean? 🙂
18:41:07 <frosch> i never used lisp or haskell languages. i do not even remember into which category scheme belongs 🙂
18:41:36 <frosch> i purely went by: what would be the most unreeadable when generated with jinja templates
18:42:34 <DorpsGek> - Update: Translations from eints (by translators)
18:43:44 <frosch> i am trying to find a list of available languages for wasm, but i only find the most stupid ones
18:44:49 <Eddi|zuHause> i think those are a vast majority :p
18:45:50 <jfs-> time to roll a new one? domain specific for OTTD
18:46:04 <jfs-> like QuakeC for is domain specific for quake 1
18:48:22 <TrueBrain> frosch: anything that has a clang-frontend, rust and typescript, basically
18:48:36 <TrueBrain> jfs-: already done years ago, called "nail" (with extension .bolt)
18:48:40 <TrueBrain> guess what nail stands for 🙂
18:49:43 <frosch> must be years ago, today you would use .true
18:49:57 <frosch> .bolt sounds like a truelight thing
18:50:27 <TrueBrain> well, NAIL was created shortly after starting NoAI, as we weren't sure Squirrel was really a good fit
18:50:38 <TrueBrain> it was functional, fully stack-based, could do all the basics
18:51:03 <TrueBrain> but to fit the theme, it was Not Another Interpreter Language; bolt came after
18:51:54 <frosch> well, i guessed the N and L correctly, but i asumed AI was just AI 🙂
18:53:21 <jfs-> Not An Intermediate Language
19:02:06 <TrueBrain> Please don't redefine the definition 😛
19:34:12 *** Flygon has quit IRC (Quit: A toaster's basically a soldering iron designed to toast bread)
19:49:32 <Eddi|zuHause> Noob And Imbecile Language :p
19:56:16 <TrueBrain> lol, that is .. noisy 😛
19:57:05 <TrueBrain> awh, even I can't make a comment anymore .. lol .. silly
19:59:36 <TrueBrain> that really did not work as I would have expected; last time I use that button 🙂 Owh well
20:03:12 <TrueBrain> I once worked for a school, being sysop; there was always this one teacher that called me with: "THE COMPUTER SHOWED AN ERROR!" - "What error?" - "I don't know, I clicked it away!". That always goes through my head when having bugs that are no longer visible; very difficult to figure out what happened 🙂
20:06:40 *** gelignite has quit IRC (Quit: Stay safe!)
20:08:16 <TrueBrain> hmm .. 10870 (cargodist crash) is bit weird
20:08:29 <TrueBrain> guess `runtime` is zero .. but why?
20:09:01 <TrueBrain> especially as the code is crafted to avoid that (in a somewhat silly way)
20:09:39 <TrueBrain> and I don't understand enough of this 😛 But maybe a `std::min` should be used, over a `+ 1 to avoid 0`
20:10:54 <TrueBrain> owh, we changed `recalc_time` ofc .. so I guess .. that could be the issue here? 😄
20:13:36 <TrueBrain> ah, I think TallTyler forgot a line when doing that change ..
20:16:18 <TrueBrain> 90% sure I am right; if someone can confirm, I can fix it 😛
20:19:06 <glx[d]> might be already fixed
20:19:22 <glx[d]> but not properly linked in the PR
20:20:30 <petern> Probably fixed yes, I think there was a different bug that was linked.
20:24:14 <TrueBrain> silly dropbox links that expired 😛
20:26:53 <TrueBrain> cool, GitHub added an `Activity` link, which shows how things ended up in the master. That is nice 🙂
20:27:12 <TrueBrain> Finally that is clearly visible 😛
20:27:14 <TrueBrain> (hihihihih, sorry :D)
20:29:30 <TrueBrain> I would say, if you have some time, confirm it, and if it is actually fixed, close it up 🙂
20:29:41 <glx[d]> not the same "function" but the same cause
20:35:56 <glx[d]> testing before and after #10820
20:56:41 <TrueBrain> Awesome, tnx for checking 🙂
20:56:54 *** keikoz has quit IRC (Ping timeout: 480 seconds)
21:17:45 *** Wolf01 has quit IRC (Quit: Once again the world is quick to bury me.)
21:44:53 <glx[d]> hmm I have some doubts about changes in #10766
21:56:54 *** nielsm has quit IRC (Ping timeout: 480 seconds)
21:58:59 <TrueBrain> Lol, we don't do trial and error are we? 😄
21:59:33 <TrueBrain> "That fixes this bug" - looks behind, sees a storm of open bugs hiding in plain sight :p
22:00:25 <TrueBrain> Anyway, tnx, will make a fix tomorrow!
22:10:26 <TallTyler> Not trial and error, suspicion and confirmation?
22:29:33 <TallTyler> Must be nice to be better than that 😛
23:24:35 *** HerzogDeXtEr has quit IRC (Read error: Connection reset by peer)
continue to next day ⏵