IRC logs for #openttd on OFTC at 2019-02-17
⏴ go to previous day
00:00:02 <peter1138> LordAro, not even joking :-)
00:00:20 <LordAro> but maybe review/merge some existing PRs first pls
00:02:32 <peter1138> glx, it starts to make sense eventually :)
00:03:07 <glx> maybe we could close #7204 as #7234 does it in a better way
00:03:15 <Eddi|zuHause> there's some very deep logic in it
00:03:23 <peter1138> Not that much logic.
00:03:31 <glx> too much templates for me
00:03:39 <peter1138> It's just a maze of templates not helped by the strange naming conventions.
00:07:30 <peter1138> glx, I think that's the plan.
00:08:22 <glx> Samu: can you rebase #7158 and remove the useless commit ?
00:12:38 *** Thedarkb-T60 has joined #openttd
00:14:10 <glx> well remove the commit first
00:15:03 <glx> but you'll still have conflicts in toolbar_gui.cpp without this commit
00:20:31 *** Thedarkb-T60 has joined #openttd
00:21:21 <peter1138> Hmm, I guess I should squash #7234 now.
00:21:27 *** DiscordIRC has joined #openttd
00:23:35 *** Thedarkb-T60 has joined #openttd
00:27:33 *** Thedarkb1-T60 has joined #openttd
00:30:28 <pnda> What's wrong with this "Thedarkb-T60" guy
00:35:01 *** Thedarkb-T60 has joined #openttd
00:37:44 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN updated pull request #7232: Change: Don't apply forbid 90 deg turn settings to ships. https://git.io/fh7DC
00:37:54 *** snail_UES_ has joined #openttd
00:38:16 <FLHerne> LordAro: I think OPF might be worth keeping simply as documentation...
00:38:33 <peter1138> Documentation of what?
00:38:33 <FLHerne> It might not be very good, but it's obvious how it works
00:38:47 <FLHerne> Whereas YAPF is just unreadable
00:39:33 *** DiscordIRC has joined #openttd
00:39:36 <FLHerne> Of how the pathfinder API works, which it should be using, etc.
00:39:46 <peter1138> No, there is no API for that.
00:40:24 <peter1138> It's literally a switch () block for each vehicle type.
00:40:57 <glx> and a heavily templated A* algorithm :)
00:46:17 <Samu> i see glx stuff got merged
00:46:51 <glx> yes that's why I suggest you remove your last commit before trying to rebase
00:47:15 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN commented on pull request #7238: Codechange: Remove assert when trying to intersect two tile areas and… https://git.io/fh5z3
00:47:48 <glx> rebase -i ^HEAD then drop the commit
00:48:16 <glx> but you will still have conflicts to solve
00:48:53 <glx> as I touched the toolbar_gui.cpp part
00:52:17 *** DiscordIRC1 has joined #openttd
00:52:22 *** supermop_Home has joined #openttd
00:53:37 <Samu> have to see how it behaves
00:57:48 <Samu> have to drop last commit
00:58:19 <peter1138> Sometimes I wonder if this is a write-only medium.
01:00:16 <Samu> how come I don't get conflicts now?
01:00:22 <Samu> something wrong happened...
01:01:39 <glx> but you should be able to move some stuff into the switch
01:02:07 *** tokai|noir has joined #openttd
01:02:07 *** ChanServ sets mode: +v tokai|noir
01:02:43 <glx> more likely the conflict would be present if you had not removed the commit ;)
01:03:17 <glx> anyway even without conflict compilation will fail
01:05:45 <Samu> include_spectator no longer exists?
01:06:13 <m3henry> glx: Oh that one surprised me. I was expecting iter to be an iterator, not an offset xD
01:09:21 <glx> you can move the #ifdef after the if (!_networking) line and insert your code there
01:09:57 <glx> but don't forget to break at the end
01:10:38 <Samu> * @param include_spectator If true, a spectator option is included in the list.
01:11:26 <glx> and I'm the first to notice that in peter1138's PR
01:13:41 <peter1138> That just means my review was bad :/
01:15:52 <DorpsGek_II> [OpenTTD/OpenTTD] M3Henry updated pull request #7165: [core] Implement SmallVector using std::vector https://git.io/fhSz0
01:16:38 <peter1138> This is bad, I'm considering using a file-scope static variable to remember the building flags used when building a house :/
01:20:46 <glx> like a local global variable ?
01:21:19 <glx> but it's done like that in other places
01:23:27 <Samu> something's quite not right
01:23:45 <Samu> even when nothing is highlighted, it still opens the global goals window or global story book window
01:24:37 <glx> unless you are a company, it then defaults to the company
01:25:35 <glx> same happens in network games
01:25:47 <glx> still it will open client list
01:28:17 <glx> hmm btw selection was probably already wrong for network games
01:30:19 <Samu> PopupMainToolbMenu(w, widget, list, _local_company == COMPANY_SPECTATOR ? widget == WID_TN_COMPANIES ? CTMN_CLIENT_LIST : CTMN_SPECTATOR : (int)_local_company);
01:31:33 <glx> result will be the same CTMN_SPECTATOR is -4 :)
01:33:37 <Samu> CTMN_SPECTATOR is being used for Global Story Book and Global Goals
01:35:29 <glx> oh right, something to add to the comment fix then
01:36:41 <Samu> for my patch, there is no client list in single player, nothing is open anyway, lucky me
01:37:36 <Samu> CTMN_CLIENT_LIST is "selected" though
01:43:38 <DorpsGek_II> [OpenTTD/OpenTTD] SamuXarick updated pull request #7158: Add: Client setting gui.start_spectator https://git.io/fhSk4
01:46:08 <snail_UES_> what should we set as a steamer’s power in our trainsets?
01:46:11 *** drac_boy has joined #openttd
01:46:22 <snail_UES_> “continuous” power or power at drawbar?
01:47:41 * drac_boy wouldn't be sure sorry
01:48:48 *** Thedarkb-T60 has joined #openttd
01:49:14 <Samu> I think I have the endif in the wrong plac
01:49:23 <glx> why did you add the _networking test ?
01:50:18 <glx> there's already !_networking test above it
01:51:00 <Samu> ifdef network confuses the heck of it
01:51:57 <glx> and yes #endif is wrong with your change
01:52:05 <glx> but it's a useless change
01:54:40 <Samu> ah. the break prevents the second part of the code from happening
01:57:11 <Pikka> snail_UES_, probably not.
01:57:15 <DorpsGek_II> [OpenTTD/OpenTTD] SamuXarick updated pull request #7158: Add: Client setting gui.start_spectator https://git.io/fhSk4
01:57:39 <snail_UES_> Pikka: what do you mean?
01:57:56 <Pikka> probably not "continuous power" or power at drawbar. :)
01:59:01 <glx> still wrong Samu, the break must be afte #endif
01:59:26 <Pikka> probably whatever value gives a pleasing gameplay effect and slots it into whatever use you're aiming for
02:00:45 <snail_UES_> sure, but that’s more about fine tuning
02:00:53 <snail_UES_> we should have a realistic base to start from
02:00:54 <glx> just in case, because _networking is false if ENABLE_NETWORK is not defined but it still safer to prevent fall through the next case
02:02:17 <Pikka> sure, but I'm not sure either of those values make sense as TTD "power", and those stats are often hard to find for steam engines anyway. So skip the research and just make it up, imo ;)
02:04:20 <snail_UES_> yeah… well, you can estimate them by using boiler pressure, grate area, superheater surface...
02:04:27 <drac_boy> pikka not exactly that hard tho .. theres a well-reused formula on the forum for to determine a good tractive for steam :)
02:04:40 <snail_UES_> tractive effort is easier
02:05:05 <snail_UES_> there is a formula involving cylinder dimensions, boiler pressure, and driving wheel diameter
02:05:11 <snail_UES_> power is trickier to estimate
02:05:12 <Samu> so that means my company_gui.cpp is also wrong?
02:06:34 <drac_boy> mind you in usa they had this ratio system where if it added up to or close to 4.0 (or was it quoted at 3.0 .. ughh I'll have to find a source to recheck) it was balanced .. but well below or above that it became more difficult (either very slippery or too much boiler output than what can be used)
02:06:57 <Samu> I don't understand this whole ENABLE_NETWORK thing, i dont even know how to test it
02:07:42 <Samu> if there is no ENABLE_NETWORK
02:07:49 <Samu> _networking doesn't exist?
02:07:51 <drac_boy> even a reboilered 4-4-2 once (it was on the IC system) had a very low number to the point that it was often shunt into the rear of the yard and left to rot as very few dispatchers ever wanted to put it on rooster
02:08:08 <glx> _networking is defined to 0 in this case
02:08:09 <drac_boy> it was just way too slippery even for a short passenger train
02:08:22 <glx> and your company_gui.cpp is correct
02:08:54 <snail_UES_> yep, but again, that’s tractive effort
02:09:24 <snail_UES_> power is proportional to the capacity of generating steam quickly and sending it to the cylinders without too much loss
02:09:27 <glx> you can mentally check the #ifdef #endif bloc like a comment block
02:09:46 <drac_boy> snail..if the boiler had a lower pressure it would had still been the same locomotive tho :)
02:09:53 <snail_UES_> problem is, you get data for continuous power for certain engines, power at drawbar for others (at different speeds)
02:09:56 <glx> #ifdef being /* and #endif is */
02:10:17 <snail_UES_> yes, same locomotive, but weaker
02:11:11 <drac_boy> come to think about it I do have two grfs that I've unofficially modified because the author for some reason left some things broken by default (eg a railcar that barely could make 30kph with one wagon which didn't compare to the real life one) .. but what do I know :)
02:12:00 <glx> checking with the compiler implies to edit the project files, and you don't wan't to do that :)
02:12:10 <DorpsGek_II> [OpenTTD/OpenTTD] SamuXarick updated pull request #7158: Add: Client setting gui.start_spectator https://git.io/fhSk4
02:12:50 <snail_UES_> if you’re referring to my set :p that’s because that railcar was almost always used by itself
02:13:26 <drac_boy> I was talking about australia and europe but :P
02:13:38 <drac_boy> which of whats the trainset you're working on?
02:13:57 <snail_UES_> haha, well there is a railcar in the NG French set that kind of fits your description
02:14:26 *** Guest1089 has joined #openttd
02:14:29 <snail_UES_> designed in the twenties, so one of the very first attempts at building a feasible diesel engine (on narrow gauge)
02:14:56 <drac_boy> oh are you the Snail with an old fiat car as forum avatar?
02:14:58 <snail_UES_> its power follows the real-life model, so it will reach full speed if it travels alone
02:15:18 <snail_UES_> the “snail” nick here was already taken…
02:15:23 <drac_boy> didn't think that was you till now .. I've looked over that thread twice anyway :)
02:15:26 <snail_UES_> so I added my location :D
02:16:14 <drac_boy> and btw although I don't know of any grf of this yet I do know a bit about the 'family' of railcars in australia tho (doesn't help I have one semi-technical book on one of them here too heh)
02:17:22 <drac_boy> various back n forth between oversea and homemade solutions .. not surprisingly the budd railcar was a bit less than impressive but still worked service for a while nevertheless .. and of course the well proven (repowering logics aside) Comet sets
02:18:19 <drac_boy> nothing too odd about actually seeing an old road-based railcar with one mini coach trailer standing (and being dwarfed) next to an australia xpt trainset heh :)
02:18:50 <drac_boy> (talk about stopping a highspeed train to transfer a few passengers from a slowpokey railcar)
02:19:31 <drac_boy> snail anyway don't mind me ;)
02:23:28 <drac_boy> snail mind if I ask roughly how many locomotives there are in that grf now?
02:24:46 <snail_UES_> about 30 locomotives (among steam, diesel and electric)
02:25:10 <snail_UES_> and about as many railcars, DMUs, and EMUs
02:25:39 <drac_boy> ah ok..sounds like I'll probably want to try it eventually (unless something else in the grf causes no-load error) .. but thats for another topic :)
02:25:45 <snail_UES_> but there is also a parameter to reduce all that crap to 10 locos and like 6 railcars :D
02:26:34 <snail_UES_> I’m working on the standard gauge now, still in planning phase
02:32:37 <drac_boy> standard? hope that included the french 131R (and maybe the army counterpart of it although I still dunno how much french people liked them tho) :)
02:33:18 <drac_boy> a little iffy that the army version was weaker but was the most utilized one of both post-war nevertheless .. oh well such is history
02:34:57 * drac_boy would also take the original tgv too :P
02:37:21 <snail_UES_> you mean the 141R?
02:37:59 <drac_boy> oh didn't noticed I hit wrong numkey .. yes the 141 :)
02:37:59 <snail_UES_> one of the most widely steamers used in the country… and american built :p
02:38:14 * drac_boy always wondered about the french system of counting per axle instead of per tires
02:42:18 <drac_boy> snail but anyway don't forget to throw in the 6700/6800 (it can be a single loco grfwise tho) .. kinda a bit of universal non-electric loco later on
02:51:17 <Samu> have been testing for hours
02:52:18 <Samu> or else I'd be running out of orders by now
02:53:27 <Samu> i just hope i dont run out of stations
02:54:52 <Samu> ah, it won't run out of stations any time soon
02:54:54 <snail_UES_> you mean the diesel BB6700?
02:55:06 <Samu> about 230 stations per ai
02:55:11 <snail_UES_> that will be in… but these days I’m too busy planning steamers :p
02:58:17 <Samu> let's see if i run out of save time
03:05:19 <drac_boy> anyway snail have fun drinking more french wines ... I'm going off sorry :P
03:08:02 *** kiwitree has joined #openttd
03:15:45 <nnyby> is the cursor style in openttd configurable? I'm sort of used to OpenGFX's nice little yellow cursor.. but i'd like to use it with the original graphiics set. this big gray error is not as nice.
05:00:44 <Eddi|zuHause> you can make a grf which just contains the arrow sprites, and use that in [newgrf-static]
06:51:25 <peter1138> snail_UES_, not drawbar power. OpenTTD already does the calculation for power varying by speed.
06:51:55 <snail_UES_> peter1138: so ideally what should I provide? continuous power?
07:09:57 <snail_UES_> anyway… looks like (boiler pressure * grate area) and superheating surface give me an R-squared of >90% on about 50 observations I collected
07:10:14 <snail_UES_> so I’m going to estimate power using those things :)
07:12:50 <peter1138> Estimate and tweak is probably bset.
07:36:04 <Eddi|zuHause> for steam engines i used a calculation i found in some old forum post by MB
08:07:32 *** andythenorth has joined #openttd
08:24:19 *** sla_ro|master has joined #openttd
08:24:59 <peter1138> When you realise you can rewrite bits to be more efficient...
08:25:09 <peter1138> First calculate the station catchment tiles.
08:25:21 <peter1138> Then Find industries and check if they are in the catchment.
08:26:08 <peter1138> Calculate station catchment tiles, and during this if you see an industry tile, you already know it's within the catchment. no need to loop all over again.
08:27:08 <peter1138> Sounds like micro-optimization, I suppose.
08:27:22 <peter1138> Should end up making the code clearer, I hope.
08:30:28 *** synchris has joined #openttd
08:36:59 <andythenorth> train stations are hard for them
08:37:28 <andythenorth> I need to add another 8-tile station to that port, for dropoff
08:38:52 <andythenorth> nowhere to put it, unless I bulldoze sea :P
08:44:04 <peter1138> Remind me, what is a port-type industry?
08:45:00 <peter1138> Isn't a port basically ... station infrastructure, not an industry?
08:45:30 <peter1138> It has storage, I suppose.
08:46:10 <andythenorth> it's just a kind of port
08:46:26 <andythenorth> the main thing is that it builds on the coast
08:46:36 <andythenorth> with water round 3 sides, approximately
08:46:45 <andythenorth> so it's limited for stations
08:46:53 <peter1138> You've missed my point :)
08:47:27 <andythenorth> stations that accept / produce cargo? o_O
08:48:30 <peter1138> Why does a port accept/produce cargo? Is it a manufacturing plant?
08:48:37 <peter1138> I suppose you could say fish->food or something.
08:48:49 <andythenorth> well there is fish -> food
08:48:58 <andythenorth> but mostly they're just a gameplay short cut
08:49:34 <andythenorth> some cargos need a source or destination that's totally at odds with RL
08:49:56 <peter1138> Can you have an industry that accepts and produces the same stuff?
08:50:17 <andythenorth> probably works technically yes
08:50:26 <andythenorth> cargo prod cb can definitely do that
08:50:27 <peter1138> And then use it as a "hub", between different companies.
08:50:37 <peter1138> one company feeds it materials from the sea
08:50:49 <peter1138> another transports that inland...
08:51:03 <andythenorth> inter-company transfers
08:51:05 <peter1138> Effectively the industry is just storage.
08:51:19 <peter1138> Except with ratings
08:52:22 <andythenorth> can nerf ratings
09:13:26 *** Progman has joined #openttd
09:46:22 <peter1138> Hmm, now I see I possibly do too much :/
09:46:27 <peter1138> I think, bike time.
09:49:33 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN commented on pull request #7235: New non-rectangular sparse station catchment area https://git.io/fh5wk
09:54:05 <Gabda> wow, there are draft PRs now
09:56:42 *** gelignite has joined #openttd
10:22:44 *** HerzogDeXtEr has joined #openttd
10:25:44 *** HerzogDeXtEr has joined #openttd
10:31:46 *** Eddi|zuHause2 has joined #openttd
10:38:56 <Gabda> peter1138: finding stations (or other objects) around tiles might be possible within O(log(#stations)) with a k-d tree structure
10:39:34 <Gabda> the main problem I found with these k-d trees that they are hard to balance, and might be easier to rebuild on station add/remove
10:41:55 <Gabda> I checked the white paper Idpl linked to my PR, and this was a data structure that might match the needs in the game
10:59:44 *** m3henry has joined #openttd
11:05:18 <DorpsGek_II> [OpenTTD/OpenTTD] M3Henry updated pull request #7165: [core] Implement SmallVector using std::vector https://git.io/fhSz0
12:07:07 <DorpsGek_II> [OpenTTD/OpenTTD] J0anJosep commented on pull request #7238: Codechange: Remove assert when trying to intersect two tile areas and… https://git.io/fh5rn
12:12:01 *** Eddi|zuHause2 is now known as Eddi|zuHause
12:16:51 <DorpsGek_II> [OpenTTD/OpenTTD] J0anJosep commented on pull request #7220: Change: Increase maximum number of orders from 64000 to ~16.7m. https://git.io/fh5rE
12:18:06 *** andythenorth is now known as Guest1130
12:18:06 *** andythenorth has joined #openttd
12:42:25 *** J0anJosep has joined #openttd
12:54:34 *** DiscordIRC has joined #openttd
13:32:47 *** Thedarkb1-T60 has joined #openttd
13:45:46 *** KenjiE20 has joined #openttd
13:53:19 *** Smokebuddie has joined #openttd
13:53:35 <Smokebuddie> Good afternoon everybody
13:54:13 <Smokebuddie> I have a issue. When i set up a server, nobody finds my game?
13:54:29 <Smokebuddie> i set the portforwarding correct
13:54:37 <Smokebuddie> and its advertised
13:55:15 <nielsm> what's your ip address?
13:56:21 <pnda> @Smokebuddie, my port forwarding is this: TCP/UDP on port 3979-3980/3979-3980
13:56:47 <Smokebuddie> 3978 & 3979 it says on the site
13:56:56 <pnda> 3979 and 3980 works for me
13:57:06 <Smokebuddie> gone try that setting on my router just a sec
13:57:51 <nielsm> you've definitely made a mistake in your port forwarding
13:58:26 <pnda> Or maybe you've set the Protocoll wrong, I remember there being options for HTTP etc.
13:59:19 <Smokebuddie> This is an example of port forwarding; in this case, UDP and TCP connections to port 3979 will need to be forwarded.
13:59:29 <Smokebuddie> this is from the OpenTTD wiki
13:59:44 <Smokebuddie> Make sure you've instructed the router to send the game data to the correct computer. If you want your server to be visible on the public server list (recommended), you will also need to forward UDP connections to port 3978.
13:59:55 <Smokebuddie> so i set my router
14:00:20 <Smokebuddie> 3979 TCP&UDP 3978 UDP
14:00:42 <pnda> my UDP is 3979 aswell, not 3978
14:02:10 <Smokebuddie> i used to be able to host but its just the last few weeks i can only join my friends. not the other way arround
14:03:13 <Smokebuddie> that's how i now end up here. in hope someone maybe knows what is blocking it
14:04:30 <Smokebuddie> @nielsm did you try to join?
14:04:37 <Smokebuddie> game name is Smokey
14:05:01 <nielsm> is what I get when I try to add the server
14:05:33 <nielsm> it can't reach on UDP port 3979 so it can't get game information
14:05:38 <Smokebuddie> its like i'm not sending any signal
14:06:00 <pnda> @Smokebuddie, try switching the UDP Port to 3979-3980
14:06:22 <Eddi|zuHause> what's your PC's internal ip? and have you set that in the router for port forwarding?
14:06:30 <Smokebuddie> @pnda i did it when you spoke of it
14:07:27 <Smokebuddie> so now i have 3979/3979 TCP&UDP, 3978/3978 UDP , 3979/3980 TCP&UDP
14:08:54 <planetmaker> Under windows its own firewall is often the culprit after you setup the port forwarding correctly on your router
14:10:16 <planetmaker> openttd will tell in console that it fails to advertize, if you start it from console and start an MP session
14:11:17 <Smokebuddie> i tried removing OpenTTD from the firewall then set the ports , start the game, start a server, accepted the firewall message to give permission. still offline to my friend....
14:11:34 <Smokebuddie> how do i start it from console?
14:11:50 <planetmaker> not a good advice... yet often when I have firewall issues with windows, I completely switch it off to see whether it is the issue
14:13:05 <planetmaker> (and often the programme then works... which then ends up with hours of swearing and weeding through MS firewall stuff)
14:25:36 <Eddi|zuHause> do you also speak with a weird austrian accent?
15:00:20 *** Extrems has joined #openttd
15:01:50 <nielsm> okay I managed to get a network game to show up on the server list, finally
15:03:14 <nielsm> had to do three things: set up port fwd in my main router (the one I like to use), set up windows firewall, and set up port fwd in my ISP-supplied router (the one I only use because it also serves as cable modem and am therefore forced to have on)
15:04:03 <planetmaker> right... there can be a 3rd one, if you have two in your home chained.
15:04:35 <nielsm> "normal" users shouldn't have that
15:07:27 <peter1138> Hmm, I wonder when the performance difference between std::set and std::unordered_set becomes noticable.
15:08:08 <nielsm> I'd expect unordered_set to be faster in almost all cases on modern hardware
15:08:58 <nielsm> (also the master server should be updated to handle the longer revision strings)
15:10:04 *** Thedarkb-T60 has joined #openttd
15:11:03 <peter1138> unordered_set may not iterate in the same order depending on compiler/machine, right?
15:11:28 <peter1138> So I proably ought to use a set when iterating otherwise... desync.
15:11:30 <nielsm> I don't think the exact hash algorithm is defined
15:11:48 <nielsm> if the iteration order matters, then yes
15:12:08 <nielsm> question is if a straight vector isn't better regardless
15:14:20 <nielsm> a std::set is a tree structure of multiple small objects, it has pretty bad memory locality properties and can easily murder cpu caches, the constant factor losses in this empirically overcome all the asymptotic runtime advantages even at very large data sizes
15:15:13 <peter1138> Mostly the sets have very few items.
15:15:23 <nielsm> even more reason to use a vector :)
15:16:21 <peter1138> I guess I was using it to be lazy with checking if an item is alreay in it.
15:17:47 <peter1138> Yeah, with vector I have to manually test every time.
15:17:56 <nielsm> I want std::stupid_set that is just a vector of items in insertion order, membership is checked by linear search, removing items from the middle moves all following items down
15:19:08 <peter1138> ^ that top answer using <algorithm> is a bit long-winded.
15:20:22 <nielsm> yeah that's the thing, "contains" is an annoying operation to express in stl on std::vector
15:33:04 * planetmaker senses small_vector's optimizations coming back
15:41:19 <peter1138> Hmm, actually if I'm iterating an unordered_set, and then inserting into a set, the order will be correct anyway.
15:49:38 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN approved pull request #7241: Fix 1585eb1a3: wrong default selection for goals and story book https://git.io/fh56a
15:53:29 <Eddi|zuHause> remember yesterday when i had trouble getting out of a cave?
15:53:37 <Eddi|zuHause> today i have trouble getting back in
15:59:18 *** DiscordIRC1 has joined #openttd
16:01:38 <Wolf01> Eddi|zuHause: take a tractor, put a drill on the front, go get that astronium in the planet core
16:02:23 <glx> I should merge an approved PR, but I'm scared
16:02:42 <Wolf01> I have troubles finding wolframite... I'm in Desolo and I only found a piece in a wreck
16:02:54 <peter1138> I think 7241 is pretty safe ;)
16:03:18 <glx> I know but we always get bug reports once merged
16:04:09 <peter1138> We need more regression tests ;)
16:04:57 <DorpsGek_II> [OpenTTD/OpenTTD] glx22 merged pull request #7241: Fix 1585eb1a3: wrong default selection for goals and story book https://git.io/fh5zi
16:05:28 <glx> yeah a game script testing the GUI is probably possible
16:06:20 <peter1138> I'm not sure it would cover this one, though.
16:06:50 <glx> indeed this one needs a spectator
16:06:59 <peter1138> When you refactor stuff and end up with a function that simply calls another function, but it's called from tons of places.
16:07:55 <Eddi|zuHause> Wolf01: not quite there yet :p
16:08:17 <Eddi|zuHause> but this arid planet is extremely annoying.. every 3 step there's a hole in the ground
16:11:19 <Samu> groups are messed up yet
16:13:07 <Eddi|zuHause> and this aluminium deposit was a total troll, not even a single complete unit
16:14:13 <LordAro> Eddi|zuHause: i just had an ammonium deposit like that, i just kept digging until i found something
16:14:15 <Gabda> Wolf, Eddi: what is this game you are talking about?
16:14:46 <Eddi|zuHause> LordAro: just use the soil refinery
16:15:21 <LordAro> Eddi|zuHause: i've not gotten that far yet :p
16:16:01 <Eddi|zuHause> LordAro: what you need ammoinum for then?
16:16:12 <Eddi|zuHause> LordAro: you can get everything for the soil refinery on the starting planet
16:16:39 <Gabda> are you playing it solo, or multi coop?
16:17:15 <Wolf01> I'm forced to play solo until they activate my new internet line, but I hope to play some coop
16:17:57 <Eddi|zuHause> i play solo, so far
16:18:19 <Eddi|zuHause> not sure i would be any use in coop with the nonexistent performance :p
16:18:20 <Samu> when I clone a vehicle, do I also clone its group?
16:19:03 <LordAro> Eddi|zuHause: because i'm still in the "pick up everything i see" stage
16:19:56 <Eddi|zuHause> i'd like a digging tool mod that increases the distance where it detects deposits underground
16:20:52 <Wolf01> I'd like to be able to deactivate mods without moving them on the backpack :|
16:21:39 <Wolf01> Like you do with the working light or the generator if placed on the shoulder attachments
16:23:09 <Eddi|zuHause> i don't really use generator
16:26:14 <Samu> damn it, why are groups failing, why do I always fail
16:26:43 <Eddi|zuHause> in general, i kinda hate single-use items (energy cells, oxygen filters, organic fuel)
16:26:52 <Wolf01> I do only in the early game, as soon as I explore I find solar panels or wind turbines everywhere
16:27:29 <Wolf01> I use energy cells on the buggy, one grants you like 40 minutes of power
16:27:42 <Wolf01> And you can find them almost everywhere
16:28:01 <Eddi|zuHause> i've not run out of power with a wind turbine
16:28:37 <Samu> openttd group sorter is really weird
16:28:55 <Samu> groups numbered over 1,000 come first than a group numbered 1
16:29:01 <peter1138> It needs to maintain heirarchy.
16:29:23 <peter1138> Do we ever sort numerically rather than alphabetically?
16:30:33 <Wolf01> I would like to reach Vesania to find some lithium, so I can start to flood the base with batteries... until I'll be able to craft some RTG
16:32:33 <Samu> I dont understand why i still have empty groups
16:32:41 <pnda> When I see "Graphics frame rate". Will there ever be a more optimized performance ingame? I mean, you don't need more than 60fps in a game like this, but if games would be more stable and load faster, when there are a lot of trains etc. running, that would be great, and would fit into that window aswell.
16:32:41 <DorpsGek_II> [OpenTTD/OpenTTD] GabdaZM opened pull request #7242: Codechange: Improve performance of town name refresh on viewports https://git.io/fh5iB
16:32:48 <Samu> i thought I would delete them all
16:34:36 <pnda> Also, is that Frame rate window included in the 1.9.0-beta2 you can download of the site?
16:35:42 <Gabda> did any of you find that scrolling if lagging on really large maps with lots of towns, or is it just me, because I use extreme number of towns (30k)?
16:36:16 <Gabda> (I only use that map for testing, not playing)
16:37:52 <Gabda> pnda: I am working on small performance upgrades :)
16:38:43 <pnda> I hope so. I think it would really benefit servers, and slower systems would maybe be able to play on larger maps
16:39:28 <pnda> I feel like OpenTTD should get Multi Threaded Support, for those with really fast PCs (like me)
16:39:44 <Eddi|zuHause> pnda: the game itself runs at 30fps, but people have been requesting things like UI to use higher frame rate
16:39:48 <Gabda> well, performance upgrades usually goes with increased memory usage, so old systems might have problem with that as well
16:40:01 <Eddi|zuHause> i never remember
16:40:28 <pnda> I mean, you really don't need more than 60fps in this type of game.
16:41:08 <Eddi|zuHause> pnda: the people using 144fps monitors seem to disagree
16:41:21 <pnda> I have a 144Hz monitor myself
16:41:54 *** Wormnest has joined #openttd
16:47:53 <pnda> Small request: When loading a game/creating a map/saving a game, could it be changed so that OpenTTD uses all available std::threads? (I have not much clue of C++)
16:48:57 <Eddi|zuHause> great. and now i glitched through the terrain into a totally different cave
16:51:24 <nielsm> pnda yes multithreading is possible "in general", but the specific structure of the save/load code in OTTD makes it unlikely to be threadable without a gigantic project
16:53:37 <nielsm> actually what do you even mean by multithreading it? run it in the background while you keep playing the game?
16:53:57 <nielsm> or split the job up so multiple parts of the loading or saving are handled simultaneously?
16:54:39 <DorpsGek_II> [OpenTTD/OpenTTD] GabdaZM commented on pull request #7120: Codechange: Improve performance of closest town lookups with cache https://git.io/fh5iF
16:56:14 <Samu> i'm getting helicopters in a group that was supposedly for send to depot only
16:56:36 <Samu> something not working, I'm sad
16:57:04 <Eddi|zuHause> that's not really relevant
16:57:44 <nielsm> OTTD is single-threaded by design, the single-threaded-ness is pervasive through the entire game, you wouldn't be able to make the simulation itself multithreaded in a reasonable manner
16:58:21 <nielsm> the reason is that the simulation needs to be predictable, and splitting it up across threads makes it unpredictable
16:58:26 <Eddi|zuHause> that just means it's shuffling around the maxed out thread between different CPUs, which is actually a terrible idea
16:58:56 <LordAro> (this is quite common amongst a lot of games - not everything is parallelisable)
16:59:04 <Eddi|zuHause> because that inhibits processor features like overclocking a single core while the others are idle
16:59:51 <Samu> i wanted AIs to multithread
16:59:57 <nielsm> there are some parts of OTTD that run on separate threads, for example graphics output may be separate depending on driver, music and sound may run in a thread, and the link graph updates (relates to cargodist) runs on a separate thread
17:00:08 <pnda> You can overclock a single processor core like you want, that's not dependant on how a program is programmed to work
17:00:35 <pnda> uhhh...My CPU only has 6 Threads, and the average CPU has 4 Threads...Over what does the game support, 16 AIs?
17:01:01 <nielsm> the cargodist link graph update is the only part of game logic that runs in a thread, and that's only possible by letting it have a deadline measured in simulated game ticks, and only integrating the thread result in game logic after the deadline has passed regardless of whether the thread finished sooner
17:01:12 <Samu> whichever AI completes the job first, gets in the queue first
17:01:18 <dwfreed> pnda: CPUs are capable of running more than 1 thread at a time
17:01:36 <dwfreed> (not really at a time, but that's what task switching takes care of)
17:01:47 <pnda> dwfreed you mean SMT/HT?
17:01:57 <LordAro> pnda: careful, you're at risk of revealing a lack of knowledge about multithreading :)
17:02:05 <DorpsGek_II> [OpenTTD/OpenTTD] GabdaZM commented on pull request #7120: Codechange: Improve performance of closest town lookups with cache https://git.io/fh5ix
17:02:12 <dwfreed> I mean the fact that CPUs aren't limited to running 1 process forever
17:02:12 <pnda> what do you mean by that
17:02:25 <milek7> for saving, maybe it is possible to fork() and finish saving in second process
17:02:34 <dwfreed> otherwise single-core single-thread processors would never have done much of anything useful
17:02:50 <Eddi|zuHause> <pnda> uhhh...My CPU only has 6 Threads, and the average CPU has 4 Threads...Over what does the game support, 16 AIs? <-- the AI threads are only threads to have local data access, they can't actually run in parallel
17:03:07 <LordAro> pnda: worrying about having more threads than CPUs is very close to nonsense
17:03:11 <nielsm> uh can anyone help me with some window system stuff? the "fill" parameter in UpdateWidgetSize, what does that mean, "fill step of the widget" means nothing to me
17:03:52 <glx> saving is already threaded
17:03:55 <pnda> LordAro: "having more threads than CPUs", what?
17:04:07 <glx> well the writing to disk part is threaded
17:04:33 <Eddi|zuHause> <milek7> for saving, maybe it is possible to fork() and finish saving in second process <-- it already does that for compressing/writing to disk
17:05:24 <glx> pnda: one CPU can have run many threads, it's called multitasking
17:06:11 <Eddi|zuHause> <pnda> uhhh...My CPU only has 6 Threads, and the average CPU has 4 Threads...Over what does the game support, 16 AIs? <-- this is total nonsense, your CPU already runs hundreds of threads before you even get to starting the game
17:06:28 <pnda> glx: They get switched arround constantly, they aren't actually running at the same time exactly.
17:06:50 <dwfreed> that's what we've been saying :P
17:07:15 <Samu> my idea was... make all AIs execute their code simultaneously, on different threads and once they need to queue commands, whichever comes first, gets it queued first
17:07:32 <Samu> then openttd runs the queue
17:07:37 <pnda> I never doubted that CPUs can just do whatever they want, and do thousands of stuff at exactly the same time.
17:07:42 <glx> I think on windows there is at almost 10 threads for openttd only
17:08:37 <dwfreed> I have threading enabled in nginx on this system; 1 master process, 2 worker process, each worker process has another 32 threads for i/o work
17:09:03 <dwfreed> s/process, e/processes, e/
17:11:01 <Samu> i just wonder if it's doable
17:11:05 <pnda> I think we were talking about different threads. But I am completely confused just now
17:11:17 <Samu> i mean, commands are already sent to a queue
17:11:20 <nielsm> pnda everyone is talking about OS threads
17:11:32 <nielsm> not (virtual or real) cpu cores
17:11:44 <pnda> yeah, I meant real cores
17:12:30 <Samu> this game is real time, so it needs to be sync'ed
17:12:46 <Samu> that's usually the task of 1 thread
17:12:50 <nielsm> Samu that's the wrong explanation
17:14:00 <nielsm> the simulation logic is single-threaded because it needs to reproduce exact results when a tick is simulated from the same starting point, that's the basis for multiplayer working at all
17:14:13 <glx> that's also why we don't use floats
17:14:31 <glx> because roundings can differ depending on the used CPU
17:15:05 <Eddi|zuHause> it's not bugs...
17:15:10 <nielsm> multithreading would either mean unpredictable results, or that everything needs to synchronize all the time losing any benefit made from parallelism
17:15:13 <Eddi|zuHause> it's undefined behaviour
17:15:49 <dwfreed> nielsm: there's still a benefit to parallelism if you have multiple simulations that need to be made every tick (eg multiple AIs)
17:15:57 <pnda> So why did my installation of OpenTTD use all 6 physical cores of mine while loading, and not just one?
17:16:33 <pnda> Probably AIs could really benefit of multiple simultaneous processes
17:16:33 <dwfreed> time(A) + time(B) becomes max(time(A), time(B)), where time(X) is the time it takes for X simulation to run
17:16:52 <nielsm> for example x87 floating point registers are 80 bit extended precision, floating point code compiled for that can have more internal precision causing it to produce different results to the same code compiled for an ARM architecture which works with 64 bit double precision all the way
17:17:36 <glx> because openttd uses many threads, even if the game core is single threaded
17:18:53 <pnda> And why does, even when creating a new 4kx4k map, my CPU not go over 35% usage?
17:18:56 <glx> when pausing during the starting newgrf scan I have 12 threads
17:19:17 <nielsm> uint fill_y; ///< Vertical fill stepsize (from initial size, \c 0 means not resizable).
17:19:22 <dwfreed> pnda: 35% out of 100% or 35% out of (numThreads * 100)% ?
17:19:49 <pnda> numThreads, OS threads or the real threads of my CPU?
17:20:05 <dwfreed> threads of the CPU, including SMT threads if available
17:20:06 <glx> nielsm: empty widget space I think
17:20:15 <pnda> Uhh, that would be 210 then
17:20:52 <dwfreed> I'm asking what the scale of the 35% is
17:20:54 <milek7> about floats, ieee754 is well-defined, right? and intermediate precision could be set with compiler flags
17:21:14 <dwfreed> milek7: CPU bugs, see Samu's mention of the Pentium bug
17:21:21 <nielsm> glx, any suggestion to a file I could read to help understanding the resizing logic?
17:21:28 <pnda> Uhh... It's 35% over all cores, idk what you mean
17:22:20 <milek7> i hope nobody uses old pentium now ;p
17:22:27 <dwfreed> pnda: so 35% out of a possible 600%, not 35% out of a possible 100%
17:23:01 <dwfreed> because that is an important distinction
17:23:29 <dwfreed> milek7: the point is that that bug is still a possibility today
17:23:56 <LordAro> with 6 cores, 35% of 600% is an issue, 35% of 100% is not
17:24:47 <LordAro> milek7: also, integer maths is generally much faster and easier to reason abouty
17:25:10 <LordAro> (35% (of 100%) is approximately 1/3, so 2 cores at 100%)
17:26:48 <pnda> So why is that bad, 35% out of 600%
17:27:39 <dwfreed> because there shouldn't be any reason it's not at at least 100% out of 600%, especially if it is a long running process
17:27:48 <pnda> And over Task Manager, it looked like none of the cores itself were over 40%
17:28:17 <LordAro> task manager isn't that great of a measuring tool
17:28:54 <dwfreed> yeah, I wish the graph per thread on task manager was vertically arranged, not horizontally
17:29:29 <dwfreed> also, in the process list on Windows 7 Task Manager, percentages are out of 100%; no idea if that holds for newer Windows versions
17:29:47 <glx> nielsm: the comment may be wrong if compared to NWidgetResizeBase::SetFill() widget.cpp:839
17:29:48 <pnda> Yeah, that seems to still be the case
17:30:11 <dwfreed> so your 35% is actually 210% then :P
17:30:33 <pnda> *This is probably irrelevant, but my GPU doesn't even hit 30%*
17:30:49 <dwfreed> which is probably fine, depending on how many threads are running map generation
17:31:38 <dwfreed> pnda: not surprising, openttd is not exactly the definition of a graphically intensive game
17:31:45 <Samu> my cpu is running 1617 threads atm
17:32:26 <planetmaker> indeed. cpu... is hardly used by OpenTTD.
17:33:09 <LordAro> planetmaker: you meant gpu, right?
17:33:57 <pnda> And a 8150 can do 3.6 million threads per second
17:35:17 <pnda> yeah just noticed myself
17:35:26 <Samu> back to where I was... my AI is bugged
17:35:36 <Samu> i fail at groups,yada yada
17:36:18 <glx> using master (ie 1.9) api or compat one ?
17:40:36 <glx> just checked the code, cloned vehicles go in the same group
17:42:17 <Samu> my road vehicles groups look fine
17:42:26 <Samu> but let me recheck, wanna make sure
17:42:48 *** m3henry has joined #openttd
17:42:51 <glx> be sure to put a vehicle in a group before starting to clone it
17:43:47 <glx> do you ever put vehicles in these groups after you created them ?
17:44:36 *** snail_UES_ has joined #openttd
17:45:15 <glx> new (non cloned) vehicles don't go in groups when built, you have to move them
17:46:27 <Samu> let me upload my stuff for u to see
17:46:46 <Eddi|zuHause> <pnda> And over Task Manager, it looked like none of the cores itself were over 40% <-- because windows is stupid and goes "oh look, CPU 1 is at 100% while the other does nothing, let's move this big thread to CPU 2", and a fraction of a second later goes "oh look, CPU 2 is at 100% while the other does nothing, let's move this big thread to CPU 3"
17:47:38 <Eddi|zuHause> so if you average that out over a second or so, each CPU is at 25% (in case of 4 CPUs and one big thread running at 100% of one CPU)
17:48:23 <planetmaker> such behaviour sounds like a good plan for cpu energy management ;)
17:49:05 <Eddi|zuHause> if by "good" you mean "ruining" :p
17:50:50 <Samu> I don't understand that spacing
17:51:01 <Samu> i see it fine on notepad++
17:51:01 <Eddi|zuHause> it means you're constantly switching power states of each CPU, instead of being able to use the boost features of a single cpu
17:51:16 <Eddi|zuHause> and you're having context switches and cache losses all over the place
17:51:36 <dwfreed> Samu: line 12 is 4 spaces, line 13 is 1 tab
17:51:46 <glx> notepad++ has a nice feature for that
17:52:45 <pnda> I like 4 spaces = 1 tab the most, default in my text editor
17:52:53 <glx> in edit menu, theres a "spaces treatments" or something like that
17:52:57 <nielsm> hmm I wonder if that makes a difference for anything?
17:53:04 <nielsm> one of the thread tasks you can indicate is Games
17:53:09 <Samu> line 185 is where i make vehicles
17:53:15 <glx> use the latest option to transform spaces in tabs
17:54:26 <glx> please convert spaces to tabs in your files and commit that
17:55:07 <LordAro> (or tabs to spaces, but consistency is better)
17:56:07 <glx> tabs is better, everyone can then use his prefered tab size
17:56:59 <LordAro> i don't disagree, but the programming world has generally chosen spaces
17:57:46 <glx> oh another useful notepad++ option in the same sub menu is "remove spaces at end of lines"
17:59:43 <glx> but this one is less important when reading ;)
18:01:58 <Wolf01> Hmmm, extenders seem to not work :/
18:02:44 <Samu> wow looks better, I didn't know there was this kind of magic in notepad
18:02:57 <Wolf01> Oh, ok, they changed completely how extenders are placed
18:03:32 <Eddi|zuHause> i haven't tried extenders... i found some, but i have no real idea why to use them
18:04:29 <glx> seems you correctly move vehicle to m_group once built
18:07:02 <glx> so the problem maybe m_group itself
18:14:45 <Samu> must identify the problem yet
18:14:54 <Samu> I'm not so sure why I have empty groups
18:15:43 <Samu> m_group should be deleted if the route is empty
18:17:11 *** DiscordIRC has joined #openttd
18:19:10 <glx> but created groups never use by vehicles are not removed it seems as your code only use vehicle list
18:20:14 <Samu> oh im not saving m_group :(
18:20:35 <Samu> no wonder i was seeing weird groups after loading
18:22:34 <DorpsGek_II> [OpenTTD/OpenTTD] nielsmh commented on pull request #7227: Replace Window::scrolling_scrollbar with Window::active_widget https://git.io/fh5Xz
18:23:37 <Eddi|zuHause> i also need an only-add and only-remove inhibitor
18:25:41 <Samu> function GroupVehicles() { this thing needs either removal, or a rework
18:26:41 <Wolf01> You can couple the inhibitor with the analyzer to only touch a particular colour of terrain
18:28:25 <Wolf01> I also found the packager really usefull, it helps a lot when you are far away and find something bigger which still function, or research items
18:28:38 <Samu> I don't need share orders to be done in that part
18:28:46 <Samu> i think it is safe to remove
18:36:59 <Wolf01> So, it's always a pleasure to fall with the entire train in a crater...
18:38:49 <Samu> as for the aircraft groups
18:39:10 <Samu> vehicles in the send_to_depot group[0]
18:39:16 <Samu> that are not heading to depots
18:39:41 <Samu> it is being used as a group for several air routes
18:40:29 <DorpsGek_II> [OpenTTD/OpenTTD] M3Henry updated pull request #7165: [core] Implement SmallVector using std::vector https://git.io/fhSz0
18:40:51 *** andythenorth has joined #openttd
18:41:05 <m3henry> The low hanging fruit is gone
18:45:08 <Eddi|zuHause> these caves in the arid planet are surely impressive, but i have no clue how to get up or down there
18:45:35 <Wolf01> Just dig and make ramps
18:46:26 <Eddi|zuHause> Wolf01: that would make sense if i'm near the wall of the cave, but looks like i'm more like on the roof of the cave
18:46:29 <Wolf01> Also consider to mod the terrain tool with a better drill mod
18:46:46 <Eddi|zuHause> i have the 1st drill mod
18:47:28 <Wolf01> Oh, I make ramps everywhere, you can see my ramps in the middle of a big cavern reaching for the ceiling
18:47:48 <Wolf01> Use ctrl to keep the same inclination and go on :P
18:48:07 <Eddi|zuHause> whenever i try that, i tend to fall off the sides to my death
18:48:15 <Eddi|zuHause> because it goes slipping
18:48:34 <Eddi|zuHause> i already got the wide mod :p
18:49:04 <Wolf01> No, really larger, you can paint them how you want
18:49:25 <Eddi|zuHause> i don't have 12 canisters with me :p
18:49:50 <Wolf01> Maybe it's because I have like 130 hours on it, but I'm not dying of falling damage since.. a lot of time
18:51:54 <Eddi|zuHause> on this arid planet i drop every 3 steps
18:52:22 <Eddi|zuHause> usually i reload because i can't get to the body to retrieve things
18:59:16 *** qwebirc49917 has joined #openttd
19:02:12 <Samu> so, when will locks become 1-tile wide only? :p
19:06:41 <Samu> how to detect wether a vehicle is already heading to a depot?
19:07:16 <Samu> is there a way that doesn't involve groups, or lists
19:07:24 <nielsm> get current order and check for OF_STOP_IN_DEPOT flag?
19:08:05 <nielsm> or well I guess you can't check the current order? only orders on the schedule?
19:09:50 <Samu> all this trouble is because I don't know of a better way
19:10:10 <nielsm> that is the way you would do it outside squirrel
19:14:29 <Samu> if (AIGroup.IsValidGroup(m_group)) {
19:14:29 <Samu> AIGroup.DeleteGroup(m_group);
19:18:45 *** gelignite has joined #openttd
19:20:16 <peter1138> Okay, found my bug. A simple negated test :p
19:21:05 <peter1138> Station Wrunnpool Pit removed from town Wrunnpool
19:21:05 <peter1138> Town at 43x12 (Wrunnpool) lost 1 stations
19:21:37 <peter1138> Just saw Alita: Battle Angel. It's pretty... Hmm... yeah, no.
19:21:47 <peter1138> Samu, no, I'm working on catchment area stuff.
19:22:10 <LordAro> peter1138: it looked rather dubious from the trailer
19:23:12 * peter1138 checks the scrollback. Persistence, as it were.
19:24:53 <peter1138> Ok, I can't be arsed with that.
19:25:44 <peter1138> Turns out my test was negated so it removed all the stations, hehe.
19:26:23 <Wolf01> <peter1138> Just saw Alita: Battle Angel. It's pretty... Hmm... yeah, no. <- I could have seen it yesterday, but it didn't catch me
19:27:08 <peter1138> Whatever happened to bridge length limits?
19:27:48 <peter1138> I thought the wooden bridge has a short limit too
19:27:51 <nielsm> hm is there a better way to detect if any AI or GS are active at all than checking each company?
19:28:22 <Eddi|zuHause> no, wooden was always somewhat infinite
19:28:57 <Eddi|zuHause> nielsm: wasn't there a command for that?
19:30:54 <nielsm> Eddi|zuHause: the context is framerate measuring and whether to do the total sum of all scripts
19:31:55 <glx> I don't think there's another way
19:32:29 <glx> as AI instance is local to the company
19:34:49 *** andythenorth has left #openttd
19:35:07 <glx> but AI:GameLoop seems to be a good place to plug your stuff
19:36:10 <nielsm> I already am doing that for individual script measurements
19:36:12 <peter1138> Problem with using caches is keeping them correct :/
19:36:34 <nielsm> the issue is making a sum of everything, and making sure that sum goes away when all scripts also go away
19:37:21 <peter1138> And basically edge cases, like a town shrinking. Is that still a thing...
19:38:39 <glx> removing the tile under the town name ?
19:39:15 <peter1138> Ooh, I should try that.
19:47:05 <Samu> this thing isn't removing correctly
19:48:02 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN commented on pull request #7227: Replace Window::scrolling_scrollbar with Window::active_widget https://git.io/fh5MB
19:48:30 <Samu> gonna make use of the removal scheduller
19:49:54 <Pikka> not sure. It might have had something to do with drawing "detailed" junctions
19:50:39 <DorpsGek_II> [OpenTTD/OpenTTD] James103 commented on pull request #7220: Change: Increase maximum number of orders from 64000 to ~16.7m. https://git.io/fh5Mz
19:51:06 <peter1138> Yes, the idea is you have to draw every combination instead the game combining bits.
19:51:16 <peter1138> But the commit date is 2013...
19:51:24 <peter1138> I don't think anyone tested it :D
19:51:38 <peter1138> I can't remember who wanted, but I suspect it was based on that that you mentioned.
19:52:08 <peter1138> 6 years ago... Maybe I should check IRC logs :D
19:53:29 <peter1138> Look at me being all sarky with Eddi|zuHause still :D
19:53:51 <peter1138> 13:01 <@peter1138> pikka, make a test grf!
19:54:45 <Pikka> I guess I got over the idea :P
19:54:59 <peter1138> It's a bit random, even then.
19:55:52 <Eddi|zuHause> i probably said something about how that's not exactly what i wanted :p
19:56:14 <peter1138> You requested a new var.
19:56:28 <peter1138> peter1138: plus a 40+ variable that gives you x%2 + y%2 << 1 so you can vary the 3rd rail for left/right side and stuff
19:57:00 <Eddi|zuHause> oh yeah, that too :p
19:57:28 <Eddi|zuHause> "if we're gonna add railtype vars, might as well add all of them :p"
19:57:48 <peter1138> Yeah... it wasn't a railtype var ;)
20:05:40 <Samu> glx, just solved the road vehicle groups being empty!
20:06:23 <Samu> different kind of problem
20:07:25 <peter1138> Hmm, AIs use distant join?
20:08:11 <peter1138> Guess they do. Neat.
20:09:28 <Samu> well, it was in the original LuDiAI
20:10:25 <peter1138> Which one is yours again?
20:12:05 <peter1138> Oh, LuDiAI crashes :/
20:12:22 <peter1138> Oh, no, it was the gamescript. Oops?
20:14:12 <Samu> AIAI also does it, but it misuses it
20:15:05 <Samu> yes, the distant join stations aren't connected to roads
20:15:24 <Samu> unconnected, bad for the looks
20:17:04 <peter1138> Ah okay, I was using that one :-)
20:20:25 <Eddi|zuHause> yay, my tractor fell through the perfectly flat terrain
20:27:14 <peter1138> Hmm, so that's the catchment area of all stations that cover that town.
20:34:05 <peter1138> Funny though, attaching a station to a large airport doesn't really extend the catchment area at all.
20:39:10 <Samu> oh boy, looking at my aircraft group management is scary
20:39:17 <Samu> groups suddenly poof and reappear
20:40:35 <Samu> I ned to think, what could possibly cause an aircraft to be in the wrong group
20:41:21 <Samu> i had aircraft in send_to_depot[0] group that weren't heading to depot
20:41:53 <Samu> the group was being misused for several air routes
20:43:19 <Samu> too bad I can't make it happen when I want
20:51:46 <peter1138> Kinda useful visualation.
20:51:56 <peter1138> Well, more useful if it only showed your own stations.
21:00:12 <Eddi|zuHause> what happened to that tile highlighting patch?
21:01:42 *** DiscordIRC has joined #openttd
21:05:49 <Samu> it always wants a commit message...
21:11:35 <peter1138> Just debugging for my non-rectangular catchment
21:16:48 <DorpsGek_II> [OpenTTD/OpenTTD] nielsmh commented on pull request #7227: Replace Window::scrolling_scrollbar with Window::active_widget https://git.io/fh5Dx
21:24:04 *** Wormnest has joined #openttd
21:28:23 <peter1138> LordAro, "tiny" bit of tiffin :o
21:28:37 <LordAro> peter1138: yes please
21:28:57 <LordAro> i just had a not-insignificant slice of victoria sponge
21:38:36 <peter1138> There's a certain theme to my instagram :p
21:40:03 *** andythenorth has joined #openttd
21:41:12 <peter1138> So, newgrf presets on bananas? :p
21:42:10 <Samu> can't make it unavailable for 1.9.0
21:42:21 <peter1138> I don't think that link will work for anyone else.
21:42:56 <Samu> maximum openttd version is 1.6.0, I need to input 1.8.0
21:43:22 <peter1138> I think it doesn't contain the last few releases either.
21:43:51 <peter1138> I'm guessing based on Samu's 'description' though.
21:44:26 <peter1138> Samu, if OpenTTD versions are missing from bananas, what you could do is actually report a bug against it.
21:44:40 <Samu> i dont know how to report banana bug
21:45:05 <peter1138> It doesn't seem to be on github, so you could try the website project.
21:45:21 <peter1138> Just make it clear you are talking about banans,
21:45:31 <peter1138> And make it clear what the problem is, but that's normal.
21:45:50 <peter1138> Yes, you could attach that picture to your bug report, that would be useful.
21:46:17 <glx> bananas is still on the old infrastructure anyway
21:46:44 <glx> and I think there were discussion about rewriting it
21:47:48 <glx> bananas issues can go in openttd issues
21:55:15 <Samu> giving 2 more days for my AI testings
21:59:46 <peter1138> And dessert is... sliced kiwi, sliced half apple, and 1/4 tub of creme fraiche
22:00:37 <peter1138> Might have room for beese and chiscuits.
22:07:08 *** supermop_Home_ has joined #openttd
22:13:12 <andythenorth> Silicon Valley GS
22:33:44 <peter1138> nielsm, hmm, "captured_widget" maybe
22:50:51 *** supermop_Home_ has quit IRC
23:17:24 *** Flygon_ has joined #openttd
23:32:42 *** supermop_Home has joined #openttd
23:54:45 <peter1138> Mmm, beese and chiscuits.
continue to next day ⏵
