IRC logs for #openttd on OFTC at 2018-01-13
⏴ go to previous day
00:00:27 <Samu> that's also what makes more sense to me
00:02:24 <Samu> a dark green for me would mean I'm using too many buses?
00:02:37 <Samu> a dark reed would mean I'm using too little buses
00:04:01 <LordAro> you're thinking in terms of vehicles rather than passengers waiting
00:06:52 <Samu> the best color is then light green, optimal usage
00:07:21 <Samu> but i see it as too many buses, i could transport them all with less buses
00:07:47 <Samu> that would make the green color a bit darker
00:12:48 <Samu> it all points to dark green being the best bus/passenger ratio
00:17:48 <Samu> let me look at AroAI screenshot again
00:19:55 <Samu> yellow and dark green are thus the best ratios
00:20:12 <Samu> it's just too constrastful these colours though
00:21:17 <Samu> it's like they mean totally opposites of each other, while in fact, they're just next to each other
00:30:50 *** Thedarkb1 has joined #openttd
00:40:19 *** Biolunar has joined #openttd
01:08:35 <Samu> 75% life gone in a combo = BALANCE!
01:16:31 <ST2> maybe here people sleep - well, we have 2 players that decided start 2rd WW - so, chats are not dead xD
01:27:43 *** TigerbotHesh has joined #openttd
02:06:21 <Rybeast> Why might my tram have atopped moving? It's just sat there. I've deleted track and asked it to 'skip orders' but nothing is happening
02:06:49 <Eddi|zuHause> it's the end of the line and no turning loop?
02:06:59 <Eddi|zuHause> or you have ridiculous timetable wait time
02:07:21 <Rybeast> No, the track is all lined up
02:07:48 <Rybeast> I don't think I've timetabled any stoppage time. ITt isn't at a station!
02:08:26 <Eddi|zuHause> try to hit the reverse button
02:08:57 <Rybeast> may have to use the cheat
02:09:18 <Eddi|zuHause> still think it's missing track
02:09:44 <Rybeast> but all the other trams are travelling where this tram is stuck
02:16:58 <Eddi|zuHause> funny memes library?
02:23:39 *** blathijs has joined #openttd
02:25:21 *** Thedarkb1 has joined #openttd
02:52:46 *** ATMunn is now known as Guest1757
02:56:29 *** ToffeeYogurtPots has quit IRC
02:57:10 *** ToffeeYogurtPots has joined #openttd
03:15:42 *** ToffeeYogurtPots has quit IRC
03:16:50 *** ToffeeYogurtPots has joined #openttd
03:19:22 *** ToffeeYogurtPots has joined #openttd
03:21:25 *** ToffeeYogurtPots has quit IRC
04:25:08 *** sim-al2 is now known as Guest1763
04:25:10 *** sim-al2 has joined #openttd
04:44:12 *** sim-al2 is now known as Guest1765
04:44:14 *** sim-al2 has joined #openttd
06:14:31 *** tokai|noir has joined #openttd
06:14:31 *** ChanServ sets mode: +v tokai|noir
07:26:00 *** HerzogDeXtEr has joined #openttd
07:48:09 *** andythenorth has joined #openttd
07:51:32 *** Celestar has joined #openttd
08:13:18 *** sla_ro|master has joined #openttd
08:50:47 *** ATMunn is now known as Guest1776
09:06:57 *** synchris has joined #openttd
10:28:01 *** FLHerne has joined #openttd
10:48:20 *** Progman has joined #openttd
10:51:37 *** ATMunn is now known as Guest1779
10:55:40 <Wolf01> Good, at least now I made it compile without crashing :P
10:56:16 *** EarthlingKira has joined #openttd
10:57:57 <EarthlingKira> Good morning, I'm new here :3
10:59:07 <EarthlingKira> OpenTTD has the nicest C++ code I've seen in an open-source game project to this day (even though I haven't looked into that many)
10:59:47 <EarthlingKira> Why is that funny?
11:00:18 <EarthlingKira> Do you think it has bad code? You always should compare relative to other experiences :D
11:00:31 <Wolf01> I've seen better code, but yes, the parts I can understand well and change are enough well done
11:00:44 <EarthlingKira> Better code in open source game projects?
11:01:01 <EarthlingKira> If so please tell me (just to have a look and maybe learn a thing or two)
11:04:12 <Wolf01> OTTD is the only open source "game" I follow, other projects I've put my hands one were designed in a different way
11:06:00 *** ATMunn_ has joined #openttd
11:06:23 <EarthlingKira> Yes, I surely also know better code in other projects which are either developed by a tight-knit group (like a company dev team) or are not game projects
11:13:56 <LordAro> EarthlingKira: OTTD's code can be pretty hairy. it's totally consistent, which is nice, but ultimately the codebase is still mostly C (given that's what it used to be) and a hell of a lot of globals
11:19:29 *** ATMunn- has joined #openttd
11:24:02 <Wolf01> Heh, I just noticed I want a "pipette" tool to activate a road building tool based on what there is on the tile I'm hovering, with NRT (and railtypes) it will come really handy
11:24:55 <Wolf01> It might be a problem on tiles with level crossings or 2 different roadtypes, but in all other cases it should be easy
11:25:20 <Eddi|zuHause> i've seen a cities syklines lets play where the guy constantly complained that there's no pipette tool
11:25:45 <Eddi|zuHause> if there are multiple types, you just cycle throught hem?
11:25:47 <Wolf01> Too much F got me used to these bad ideas :(
11:26:00 <Wolf01> Yeah, good suggestion Eddi
11:27:35 <Wolf01> Ok, now I got rid of all the occurrencies of the check for road catenary and moved all of them in a single function
11:28:04 <Wolf01> So I can change the behaviour in a single place
11:33:18 <andythenorth> I'll draw ships then
11:40:19 *** Breckett has joined #openttd
11:40:49 <Wolf01> Slow connection is slow
11:44:30 <andythenorth> do we have test grfs? o_O
11:44:41 <andythenorth> supermop: trolleybi has catenary eh?
11:45:10 <Wolf01> Branch: both-catenaries
11:45:24 <Wolf01> Yes, checking with trolleybi
11:51:39 <Borg> anyone know if.. and how.. to change GS settings via console?
11:52:02 *** Arveen2 has joined #openttd
11:53:15 <EarthlingKira> @all for players who like to make rather short cargo routes (and not be forced to make long routes, but still be able to do it) I've created balance patches:
11:53:55 <andythenorth> lose money on short routes? :o
11:54:00 * andythenorth never has that problem :)
11:54:30 <andythenorth> how short is short?
11:54:35 <EarthlingKira> I've actually not yet found a patch or NewGRF who's done it like that. I've only found a NewGRF to limit maximum payments to a fixed distance (which sucks, sometimes you have to make long routes) and I've found a NewGRF which has linear payments for every distance (this also sucks, bus stations 2 tiles away give too much profit)
11:54:42 <EarthlingKira> @andythenorth I'm talking about like 3-5 tiles
11:55:24 <EarthlingKira> Or transporting coal 10 tiles away with trucks also made me loose money. (4% inflation, high maintenance costs, infrastructure maintenance enabled)
11:55:30 <andythenorth> ok, I would think 'short' is about 10 tiles
11:55:36 <andythenorth> you have inflation on?
11:55:40 <andythenorth> inflation is a bug
11:55:57 <EarthlingKira> And inflation is coupled to loan interest says the setting description
11:56:02 <EarthlingKira> Nah, I like inflation
11:56:03 <andythenorth> if had trunk commit rights, inflation would be deleted
11:56:04 <Wolf01> andythenorth: wires are a pixel mess :D but at least you can see both poles for road and tram
11:56:32 <Wolf01> We are lucky you don't have commit rights ;)
11:56:49 <EarthlingKira> @andythenorth do you have high maintenance costs and infrastructure maintenance enabled then?
11:57:15 <andythenorth> also I play with my own newgrfs
11:57:27 <EarthlingKira> Ah well sure then you can make a profit with routes ~ 10 tiles :D
11:57:55 <EarthlingKira> I really like my small city bus lines with stations everywhere and I want them to make some money
11:58:55 <andythenorth> my bus stops are about 10-12 tiles apart
12:00:07 <EarthlingKira> What about the houses inbetween?
12:00:21 <Wolf01> I try to make bus stops so catchement areas don't overlap :P
12:00:32 <EarthlingKira> It has only 3 catchment area tiles, right?
12:00:51 <EarthlingKira> [A][a][a][a][b][b][b][B]
12:00:56 <EarthlingKira> That would be 7 tiles distance
12:00:57 <andythenorth> I was counting manhattan distance ;)
12:01:16 <EarthlingKira> are you letting your buses drive zigzag in case? :D
12:01:27 <andythenorth> just the shape of the town
12:02:12 <Eddi|zuHause> i ususlly put my stops closer
12:02:19 <andythenorth> anyway GL EarthlingKira
12:02:30 <andythenorth> all attempts to nerf distance payment have been resisted
12:02:39 *** Maraxus has joined #openttd
12:04:08 <Wolf01> So, did you try it, andythenorth?
12:04:17 <andythenorth> Wolf01: have you pushed one that draws both>?
12:04:24 <andythenorth> I've got the de-duplicated commit
12:04:27 <peter1138> Hmm, do I really want to download 739 MB just for... audio drivers?
12:04:30 <andythenorth> but I only see one catenary
12:05:10 <Wolf01> It contains both, strange
12:05:42 <Wolf01> Check road_cmd.cpp:1474
12:05:52 <andythenorth> defers to whichever was built first I think
12:06:02 *** Biolunar has joined #openttd
12:06:07 <Borg> EarthlingKira: in city transportation is usually donated from city officials dude
12:06:16 <Wolf01> It just draws road and then tram
12:06:29 <Wolf01> Tram is always on top of road one
12:06:42 <Borg> go to multiplauer server.. find mogul who asks for city grow.. and he will donate you some cash :D
12:08:22 <andythenorth> but not like that
12:08:28 <andythenorth> wonder if I have different unspooled
12:08:38 <Wolf01> Grey on grey is a bit hard to see
12:08:48 <andythenorth> I have v2 unspooled
12:08:54 <andythenorth> the poles seem to overlap
12:09:33 <Wolf01> I have.. I wonder which version :D
12:09:44 <EarthlingKira> @andythenorth Thank you, I don't think that this patch should go into trunk (although I could imagine it being based on a setting)
12:10:05 <EarthlingKira> And the patch doesn't just reduce distance payments, it rather changes the algorithm to be partially logarithmic
12:10:16 <EarthlingKira> Earn more on short distances, less on long distances
12:11:46 <andythenorth> Wolf01: can you DM me the version of unspooled you have?
12:11:56 <EarthlingKira> 2 years ago I already patched my own game and used a real logarithm but then I've seen that I can't use doubles/floats because of desync possibilities m(
12:13:13 <EarthlingKira> Oh and I have one final patch @channel
12:13:40 <EarthlingKira> I did not want a complete breakdown system overhaul like other patches do... I just wanted the chance algorithm and reduced breakdown chance settings to make more sense
12:14:11 <EarthlingKira> So this is just a small tweak to the original chance calculation algorithm which makes it *MUCH* more enjoyable and reliability actually changes things now
12:14:37 <EarthlingKira> And reduced breakdown chance no longer causes vehicles between 60% and 100% reliability to behave exactly the same...
12:15:34 <Wolf01> Meh, I have 0.005bps upload
12:16:18 <Wolf01> andythenorth: check forum PM
12:17:28 <Borg> honestly.. prospecting algoritm should be fixed.. so it will not place production and raw industries close to each other
12:17:38 <Wolf01> (I hope it was uploaded fine)
12:19:03 <andythenorth> still don't see separate poles
12:19:07 <andythenorth> but both wires are drawn
12:19:24 <Wolf01> Did you try with all different roadtypes?
12:19:30 <andythenorth> the poles are drawn
12:19:32 <andythenorth> they just overlap
12:19:41 <andythenorth> sprite inspector finds them
12:19:51 <andythenorth> ok so it probably works eh
12:20:54 <andythenorth> ok found a non-overlapping type
12:21:09 <andythenorth> so what authors need to do is
12:21:26 <andythenorth> don't try and differentiate wires
12:21:33 <andythenorth> use the poles, preferably colour-themed
12:21:51 <andythenorth> poles are much more legible + different
12:22:08 <Wolf01> Wires are a line of grey pixels on grey
12:22:11 <andythenorth> they barely show anyway because of tram rails
12:22:26 <andythenorth> so unspooled does it fine
12:22:41 <andythenorth> I reckon merge that
12:22:49 <andythenorth> it's just drawing code?
12:22:53 <andythenorth> no savegame or anything?
12:23:15 <EarthlingKira> Btw, one final question regarding my patches: How's the process in case they are trunk-worthy, who would decide that? I'm willing to create boolean options and settings with parameters to change the magnitudes :)
12:23:29 <andythenorth> cool, merge to road-and-tram-types
12:23:37 <andythenorth> EarthlingKira: currently no-one knows :)
12:23:50 <andythenorth> it's pretty much luck whether anything gets to trunk
12:24:04 <Wolf01> Are you sure? I would wait for frosch green light
12:25:20 <Wolf01> Also I would make the patch better, merge the "DrawRoadCatenary" with "DrawCatenary"
12:31:49 *** Rybeast has joined #openttd
12:32:53 <Rybeast> Any ideas why one of my trams might have completely stopped and failed to work? I've checked timetabling, pathing, when some trams push past they're able to go in both ways the broken tram should go, but now there is a huge tailback and I don't know how to rectify
12:33:23 <andythenorth> Wolf01: ok code review probably a good idea
12:33:31 <andythenorth> Rybeast: missing track piece?
12:34:49 <Rybeast> It's definitely not a missing track piece. Trams are still able to follow the exact path in all three ways
12:35:38 <Wolf01> It just stopped there with speed=0?
12:35:41 <andythenorth> red flag stopped?
12:36:00 <Rybeast> Doesn't seem to be. All the trams are green
12:36:42 <andythenorth> is it broken down?
12:37:19 <Wolf01> I ran into this problem with NRT
12:37:20 <Rybeast> i wondered if I'd set timetabling up incorrectly, but it doesn't seem to be affected by my moving and editing timetabling
12:41:40 <Wolf01> A branch we are working on
12:41:49 <Wolf01> Did you change the grf?
12:44:59 <Rybeast> well, I added some last night, but I don't think I changed any, no
12:45:05 <Rybeast> haven't edited at all
12:45:27 <Wolf01> But you changed active grfs in a running game?
12:46:37 <Rybeast> they were there before I started
12:46:43 *** frosch123 has joined #openttd
12:46:56 <Wolf01> Do you have a savegame of before the problem occours?
12:50:54 <Rybeast> I do, but it doesn't show anything there either.Everything does what it should do. It was when I added the newer trams I guess, but I don't know why
12:53:37 <Wolf01> Could you replicate it without adding the new trams?
12:54:37 <Rybeast> I'm going to have a quick look now
13:01:40 <andythenorth> Wolf01: so just town roads to solve now?
13:01:47 <andythenorth> everything else is blah blah
13:01:58 <Wolf01> And fix scenario editor
13:02:14 <Wolf01> There's already a branch for that
13:03:22 <Wolf01> But that does a lot more, so I'll need to extract a part and apply it to the base branch
13:03:52 <EarthlingKira> Btw, do you think a rework of the advertisement campaigns make sense, i.e. does anyone use them?
13:04:43 <Wolf01> I used it 2 or 3 times
13:04:56 <EarthlingKira> They actually sometimes make sense in ECS to increase production rate of factories
13:05:07 <EarthlingKira> But otherwise the effect is gone too fast
13:07:26 <EarthlingKira> I think the effect of a large advertisement campaign should stay for 3 or 6 months and have a larger radius
13:08:58 <EarthlingKira> I think the advertisement system should be balanced around the idea that you can *actually* increase your revenue with doing non-stop advertisements
13:09:11 <EarthlingKira> (when you have lots of stations in/around a certain city)
13:09:36 <andythenorth> the \\ are noticeable...
13:12:17 <frosch123> well, windows, ¯¯\o/¯¯
13:14:30 <frosch123> andythenorth: i guess your problem is bin/find-files
13:15:00 <frosch123> it thinks "\" is part of the name and no directory separator
13:15:16 <frosch123> and then escapes it
13:17:13 <frosch123> i don't think it is fixable without try-error on an actual windows system
13:17:28 <frosch123> noone can predict when those tools use / or \
13:32:55 <Wolf01> On windows you should use / anyway
13:33:12 <andythenorth> I have NFI how to even start solving this :)
13:33:34 <andythenorth> I do have windows VM, but it's vanilla, no dev tools of any kind
13:33:39 <andythenorth> just for IE testing
13:33:55 <Wolf01> I have windows but it's bloated of dev tools
13:41:16 <EarthlingKira> @Wolf01 but that is because windows dev tools are bloated by themselves. Visual Studio up to 82 GiB install size m(
13:42:00 <Wolf01> Nah, I installed even more dev tools
13:45:59 <Wolf01> I was (am) a web developer, so I also have a lot of node, ruby, sass related stuff, php stuff too, android stuff (both VS xamarin and android studio)
13:46:57 <Wolf01> And a bit of asm and plc stuff
13:47:55 <EarthlingKira> Web development on Windows :'( I'm also (partially) a web developer, but I switched to Linux 10 years ago for coding and never looked back :3
13:48:54 <Wolf01> Now I use WLS to do things I didn't could do before, It was made exactly for that, so I don't even need to dualboot
14:32:22 *** Thedarkb1 has joined #openttd
14:49:27 <Samu> hmm patches to nerf stuff!
14:57:36 <Jamesy> Hello. I've been setting my vehicles to have services every x days or when the percentage drops to y. The vehicles don't seem to be headed back to depot to be serviced however. How would I sort this?
14:58:16 <Samu> the depot might be too far away
14:58:31 <Eddi|zuHause> i think it's "percentage drops <by> X (from the maximum)"
14:58:42 <Jamesy> But they all pass the depot between the two stations?
14:59:16 <Eddi|zuHause> so if you set 80%, and the max is 93%, it would need to be 13% to trigger, or something
14:59:57 <Jamesy> so if my percentage was 60%, it could drop down to 33% before needing servicing?
14:59:58 <Samu> i think 15% is the ideal value
15:00:06 <Samu> the default of 50% is kinda ridiculous
15:01:22 <Jamesy> it is, but even then I have trains who have a reliability of 0
15:02:07 <Samu> hmm this might be something worth testing
15:02:07 <Jamesy> so i'm having to identify these trains and send t depot manually and adjust the servicing requirements
15:02:19 <Samu> hunting the ideal breakdown as a % value :p
15:03:00 <Jamesy> which makes sense, because the trains reliability can't drop any further, so is the game identifying that the train doesn't need servicing because it hasn't dropped by x in that time, even though it has dropped below the baseline already?
15:04:03 <Samu> do you have send to depot orders?
15:04:13 <Samu> they're not automatically serviced
15:06:00 <Samu> i feel like testing EarthlingKira patches
15:06:37 <Jamesy> I haven't put in 'send to depot' - if I do, is there an option to 'send if'?
15:08:59 <Samu> try "service at nearest depot" or "service at <input depot>"
15:09:00 <Eddi|zuHause> yes, there is "service at depot" if you ctrl+click while placing
15:09:26 <Samu> when the vehicle executes that order, and it needs servicing, it will head to depot
15:09:34 <Eddi|zuHause> that means "go to depot, but only if servicing is needed"
15:09:35 <Samu> if it doesn't needs servicing, it just skips the order
15:12:16 <Jamesy> Can I set out a precedent to all my vehicles of that type? Like, I've got 18 freight trains running around?
15:13:16 <Eddi|zuHause> you can do shared orders
15:13:25 <Eddi|zuHause> so all 18 freight trains have a common order list
15:13:42 <Jamesy> Ooh, how do I do that?
15:13:44 <Eddi|zuHause> then all future changes will apply to all of them simultaneously
15:14:30 <Eddi|zuHause> you assign the orders to the first train like normal, and on the second train, you don't assign any station, but with the "go to" tool you ctrl+click on the first train
15:14:53 <Eddi|zuHause> (without ctrl it makes a copy of the order list, with ctrl it shares the order list)
15:15:26 <Jamesy> excellent, so by 'control clicking' I'll only need to do the one order?
15:22:49 <EarthlingKira> @Samu Feel free, and if you have suggestions I'm glad to hear them :)
15:25:28 <andythenorth> so who wants to patch FIRS for Windows then? o_O
15:38:05 *** Montana_ has joined #openttd
15:43:16 <supermop> yo andythenorth what's this talk of poles?
15:43:24 <supermop> and do you need old versions?
15:43:35 <supermop> all versions should have same pole sprites tho
15:44:04 <Montana_> i updated a translation for FIRS this morning
15:44:32 <Montana_> is that update automatically uploaded to FIRS repo?
15:45:06 <frosch123> normally you also get a playable .grf at about 17:15 CET, but probably not today
15:46:22 <supermop> pole sprites do need some work
15:47:06 <supermop> and there is a bug with front most pole not showing on U tiles where the loop is at the lower left or right
15:47:08 <andythenorth> Montana has gone
15:47:14 <andythenorth> I am probably going to release FIRS today
15:48:41 *** Montana_ has joined #openttd
15:48:55 <Montana_> Sorry, internet connection interupted
15:49:14 <Montana_> but i did read your message Frosch
15:49:28 <frosch123> [15:47] <andythenorth> I am probably going to release FIRS today
16:01:26 *** sla_ro|master has joined #openttd
16:06:45 <Samu> testing the breakdown patch, i hope it screws aircraft
16:07:53 <EarthlingKira> @Samu My breakdown patch reduces the amount of breakdown occurrences, why should it screw aircraft?
16:08:34 <EarthlingKira> Do you mean compared to playing without breakdowns? ;-)
16:10:58 <Samu> i'm unsure what this patch does, you say ships never breakdown if they had 60%
16:11:23 <EarthlingKira> Oops, maybe my wording is misleading
16:12:34 <EarthlingKira> In the vanilla OpenTTD when playing with "reduced breakdowns" a ship with 60% reliability and a ship with 100% reliability will break exactly the same amount of time
16:13:08 <EarthlingKira> Forget my last sentence :D
16:13:37 <EarthlingKira> A ship which has >= 60% reliability (so no matter if 60%, 80% or 100%) will not profit at all from "reduced breakdown chance" setting
16:14:04 <EarthlingKira> It will break exactly the same amount of time no matter if you have the setting enabled or not
16:15:50 <EarthlingKira> With my patch you a) really always have a reduced breakdown chance when that setting is enabled and b) reliability % has a much stronger effect on breakdown chance
16:15:53 <Samu> this is a bad timing for building openttd, i'm testing AIs at the same time :(
16:17:12 <EarthlingKira> And c) when playing with "reduced breakdowns" setting a train with 60% reliability now breaks more often than one with 100%, as you would expect
16:17:55 <EarthlingKira> In vanilla openttd with "reduced breakdown" setting trains from 60% to 100% and ships from 20% to 100% reliability are indistinguishable
16:18:53 <EarthlingKira> I would really hope that that patch is merged into trunk, because I want to play with reduced breakdowns but still have a train with 60% reliability be worse than a train with 100%, right?
16:19:22 <Samu> i'm not sure if I follow you yet, but it peaked my interest
16:20:57 <EarthlingKira> Okay I explain the other way around: in vanilla OpenTTD:
16:21:12 <EarthlingKira> - ships always get a flat invisible 40% bonus to reliability
16:21:32 <Samu> if (v->type == VEH_SHIP) rel += 0x6666;
16:21:50 <EarthlingKira> - every vehicle/ship/airplane gets another flat invisible 40% bonus to reliability when "reduced breakdown" is active
16:22:32 <EarthlingKira> That means a train with 60% reliability has an effective reliability of 100% with that setting enabled, while a train with 100% reliability also has an effective reliability of 100%
16:22:38 <Samu> if (_settings_game.difficulty.vehicle_breakdowns == 1) rel += 0x6666;
16:23:37 <Samu> that is the part I still don't get it
16:23:59 <Samu> _breakdown_chance[(uint)min(rel, 0xffff) >> 10]
16:24:43 <EarthlingKira> the min(rel, 0xffff) basically says that the maximum reliability you can have is 100%, no matter how much was added before
16:25:07 <EarthlingKira> And the code around that part is to convert this big number into one that is used for the breakdown chance
16:26:44 <Samu> static const byte _breakdown_chance[64] = { 3, 3, 3, 3, 3, 3, 3, 3, 4, 4, 5, 5, 6, 6, 7, 7, 8, 8, 9, 9, 10, 10, 11, 11, 12, 13, 13, 13, 13, 14, 15, 16, 17, 19, 21, 25, 28, 31, 34, 37, 40, 44, 48, 52, 56, 60, 64, 68, 72, 80, 90, 100, 110, 120, 130, 140, 150, 170, 190, 210, 230, 250, 250, 250, };
16:26:49 <EarthlingKira> Yes that's the one
16:28:21 <EarthlingKira> you get a comparison value of 250 when you have 100% reliability and a comparison value of 3 when you have 0% reliability
16:29:19 <EarthlingKira> In vanilla OpenTTD the reliability only says how long it will take to reach the breakdown threshold
16:29:52 <EarthlingKira> So a higher reliability does not really decrease breakdown chance, it only takes longer (sounds like it's the same, but it's not :D)
16:31:49 <Borg> hmmm ... Cargo flow tool is invaluable tool now.. when using stockpiling industries :)
16:32:25 <DorpsGek> Samu: Error: 'ffff' is not a defined function.
16:32:35 <DorpsGek> Samu: Error: Something in there wasn't a valid number.
16:33:22 <Samu> whatever, my calcultator says that is = 0
16:33:39 <Samu> _breakdown_chance[0] = 3 ?
16:35:00 <Samu> 3 <= v->breakdown_chance
16:35:18 <EarthlingKira> 0xffff >> 10 = 63
16:36:03 <EarthlingKira> So yes 63 in that case :D
16:36:21 <EarthlingKira> You probably did not deactivate hexadecimal mode when entering the value for rightshift into your calculator
16:36:26 <Samu> or, right I was doing 0x10, not the real 10
16:36:51 <EarthlingKira> @calc 65535 >> 10
16:36:51 <DorpsGek> EarthlingKira: Error: Something in there wasn't a valid number.
16:37:03 <Samu> _breakdown_chance[63] = 250
16:38:39 <Samu> hmm ok, must investigate how is 'rel' calculated too
16:39:07 <EarthlingKira> I think rel is simply 100% = 0xffff and 0% = 0x0000
16:39:23 <EarthlingKira> It's the reliability
16:39:34 <Samu> if (v->type == VEH_SHIP) rel += 0x6666;
16:39:58 <Samu> if (_settings_game.difficulty.vehicle_breakdowns == 1) rel += 0x6666;
16:41:26 <EarthlingKira> I don't like vanilla algorithm. I want to have less breakdowns but at the same time I want a train with 60% reliability to break more often than one with 100%
16:41:28 <Samu> it then compares with itself without the added bonus, weird
16:42:09 <Samu> ah nvm, im seeing this wrong
16:42:31 <Samu> there's v->reliability and v->breakdown_chance, two different things
16:43:12 <EarthlingKira> Even though the naming is... well... misleading
16:43:53 <EarthlingKira> But with my patch the breakdown_chance actually acts like a real chance
16:44:04 *** Thedarkb1 has joined #openttd
16:44:47 <EarthlingKira> Personally the only thing I'd still like to add is (linearly) increased wear on vehicles based on current velocity
16:45:18 <EarthlingKira> Because when you use that system and have a "go to depot" order and have trains with 600 km/h, you have virtually no more breakdowns
16:45:56 <EarthlingKira> @Samu What's your opinion on that?
16:46:38 <EarthlingKira> (This would also fulfil your wish to screw airplanes :3)
16:46:56 <Samu> I'm still building, it's taking longer than I expected, think it's a bad time for testing this
16:47:07 <Samu> must finish my AI tests first
16:49:15 <Samu> something that lets me test 8 AIs in 8 4096x4096 maps while building openttd in visual studio
16:51:36 <EarthlingKira> Ryzen Threadripper
16:56:15 <Samu> yeah, i thought of that, but then intel did some price drops, so i'm currently undecided. also my current rig is still doing fine, except when i start testing crazy stuff
16:56:27 <Samu> not sure if it's the right moment
16:57:28 <Samu> v->reliability is a uint16
16:57:34 *** Breckett has joined #openttd
16:57:36 <Samu> v->breakdown_chance is a byte
16:58:59 <Samu> int chance = v->breakdown_chance;
16:59:14 <Samu> i wonder if chance can sometimes be a negative value? :/ i fail at programming
17:00:01 <EarthlingKira> An int can be negative if you cause it to be negative, yes ;-)
17:01:27 <EarthlingKira> No, an int has more bits than a byte
17:02:16 <Samu> ah, so if v->breakdown_chance is 255 byte, int chance is 255 too, i see
17:03:59 <Samu> i wonder what's the priority of visual studio, seems like idling...
17:04:22 <Samu> granted, I only got about ~5% cpu free
17:06:01 <Samu> (GB(r, 16, 8) <= v->breakdown_chance && GB(r, 16, 8) >= _breakdown_chance[(uint)min(rel, 0xffff) >> 10])
17:08:25 <EarthlingKira> My patch combines the element of the vanilla algorithm that a minimum breakdown amount must have been accumulated and additionally applies some chance based on reliability to see if the vehicle should really breakdown
17:08:50 <supermop> EarthlingKira: interesting passenger patch
17:09:25 <Samu> the Random() is a biiig number
17:10:53 <Samu> 1111 1111 XXXX XXXX 1111 1111 1111 1111
17:11:35 <Samu> that's a value from 0 to 255, right?
17:11:51 <EarthlingKira> @supermop Thanks :) I enjoy building PAX networks much more with that patch, small villages actually make a profit and big cities are no longer overloaded
17:12:17 <EarthlingKira> @Samu Right, this GB() thingie is just used to get smaller random numbers
17:13:15 <Samu> if the value is 0, it still breakdowns
17:15:51 <supermop> the problem is made worse by the town zone pattern too -
17:17:01 <supermop> as once a town gets very large, it is essentially primarily the densest zone by area, so the majority of houses are then producing the insane amount of passengers
17:17:19 <EarthlingKira> Yes :) You should try playing with my patch, it's much better :3
17:19:19 <Samu> my patches are ugly coded, but I like the ideas they provide ;p
17:19:59 <Samu> ah, in the forum i go by the name of xarick
17:24:26 <EarthlingKira> Regarding code "quality" I'd say that it's not perfect that the aircraft specific code is mixed into the breakdown code which increases coupling ;-)
17:27:22 <EarthlingKira> I'd be interested to write really really good code and do some refactorings and increase some game design mechanics, but for that I'd need to know that it is potentially wanted and could be merged into trunk
17:41:17 <EarthlingKira> @Samu What exactly did you mean when you said check your patches? Do you want a rating, some improvement feedback, a code review or maybe even code improvements? ;-)
17:48:19 <EarthlingKira> @supermop I've answered to your feedback, what do you think? :)
17:48:36 <Samu> it has finally built! testing the breakdown patch
17:48:59 <EarthlingKira> Are you testing with reduced breakdown rate or with normal breakdown rate setting?
17:49:32 <Samu> testing with reduced atm
17:49:51 <EarthlingKira> Okay :) I've tweaked it only for reduced setting :D
17:50:57 <EarthlingKira> normal breakdown setting still has the reliability chance effect, but I've not tested/tweaked it with this setting
17:59:43 <Samu> GB(r, 16, 8) what exactly is this doing? it's a bit confusing
18:00:47 <Samu> r 406248016 unsigned int
18:02:29 <Samu> it's randomly getting a value from 0 to 255
18:02:41 <Samu> but... what's the intention
18:05:48 <LordAro> "random" is probably the wrong word there
18:06:20 <LordAro> GB is a very well used function, i'd recommend looking it up and working out what it's doing
18:07:13 <LordAro> ah right, i thought you were referring to GB
18:07:37 <Samu> the pseudo-random number generator thing
18:10:21 <Samu> i see that v->breakdown_chance can at times stay the same
18:12:20 <Samu> it was previously always adding 1 to v->breakdown_chance every day
18:13:54 <Samu> I wonder... why do ships get special treatment :(
18:14:25 <EarthlingKira> @Samu I didn't want to change too much
18:14:38 <EarthlingKira> To differentiate from breakdown overhaul patches
18:17:01 <Samu> what will happen when reliability is 0%
18:17:19 <Samu> seems that reduced breakdowns will make them quite good, lol
18:39:36 <Samu> aha, it's finally starting to make sense for me
18:41:13 <Samu> EarthlingKira: wouldn't it make better sense to have a 0% reliability with reduced breakdowns to behave equal as a 0% reliability with normal breakdowns?
18:43:26 <Samu> just saying, in multiplayer servers, unmaintainced companies should have their old vehicles cannibalize the profits
18:48:08 <Samu> 1.7.2 - 30 breakdowns in 3 years, 0 months
18:48:25 <Samu> patch - 30 breakdowns in 4 years, 3 months
18:49:53 *** m3henry has joined #openttd
18:51:13 <EarthlingKira> @Samu That seems fair, doesm
18:51:43 <Samu> now i'm testing normal breakdowns, will take a while
18:53:36 <Samu> from 100% to 0% reliability, it broke down 45 times, in 1.7.2
18:53:56 <Samu> from 100% to 0% reliability, it broke down 24 times, in patched
18:54:37 <Samu> waiting for 100 breakdowns
18:56:04 <m3henry> I'm never sure how I feel about breakdowns
18:57:08 <Samu> it feels too forgiving now
18:57:16 <m3henry> Sometimes I feel like they add depth and make descions more interesting, but they become rather annoying when density rises
18:57:57 *** Wormnest has joined #openttd
18:58:23 <m3henry> If rolling stock pooling was a thing, then I could see it being very interesting
19:00:13 <Wolf01> frosch123, what do you think about drawing both the catenaries? Should we merge it to NRT?
19:03:09 <Samu> 1.7.2 - 100 breakdowns in 3 years, 8 months
19:03:24 <Samu> patch - 100 breakdowns in 7 years, 7 months
19:04:06 <Samu> i think 0% reliability needs special care
19:05:16 <andythenorth> breakdowns are tedious
19:05:35 <andythenorth> even on, they're trivially prevented by forced-servicing
19:06:12 <m3henry> I liked the idea of making breakdowns reduce the power output of engines when they happen
19:06:28 <m3henry> There was a patch floating around that did that
19:06:38 <m3henry> rather than a full stop
19:07:22 <frosch123> makes it more visually clear
19:09:41 <Samu> patch - 30 breakdowns in 2 years, 9 months, reduced breakdowns
19:11:46 <Samu> strange, i had the impression it was breaking down more often when the reliability was not yet 0%, but close to 0
19:12:25 <Samu> how can I quantify this :(
19:12:29 <Wolf01> I cleaned it up a bit, renamed functions too
19:15:40 <Wolf01> Cat is done, andythenorth should be happy
19:19:03 <andythenorth> now just town crap :)
19:19:24 *** Thedarkb1 has joined #openttd
19:19:51 <andythenorth> Is there a proposed implementation for town?
19:19:59 <andythenorth> there are a few ways it could be done
19:20:34 *** Stimrol has joined #openttd
19:21:05 <andythenorth> - take ownership
19:21:09 <andythenorth> - don't take ownership
19:21:26 <andythenorth> - allow build if compatible with ROAD
19:21:30 <andythenorth> - allow building any
19:22:35 <Wolf01> Also there is default road for towns
19:24:05 <Wolf01> I don't know what could happen if a grf disables standard road
19:32:20 <Samu> [18:09] <Samu> patch - 30 breakdowns in 2 years, 9 months, reduced breakdowns*
19:32:30 <Samu> i meant normal breakdowns, sorry
19:33:53 <EarthlingKira> > patch - 100 breakdowns in 7 years, 7 months
19:33:58 <EarthlingKira> Was this with reduced breakdown setting or normal?
19:34:29 <andythenorth> Wolf01: I could make a grf that tests disabling standard ROAD
19:34:33 <andythenorth> not sure how right now
19:37:14 *** chomwitt has joined #openttd
19:38:31 <Samu> i started a bus with 100% reliability, then removed depot
19:38:53 <Samu> so, the majority of breakdowns was with 0% reliability
19:39:49 <Samu> from 100% to 0% it broke down 24 times
19:40:08 <Samu> let me time how long it takes to go from 100% to 0%, brb
19:40:33 *** Stimrol has joined #openttd
19:45:09 <Samu> 2 years, 4 months, i think
19:45:25 <Samu> @calc 2 * 12 * 30 + 4 * 30
19:45:33 <Samu> yeah... seems to be that
19:53:53 <supermop> Wolf01: presumable it would just fall back to whatever gets first defined then
19:54:38 <supermop> like if you disable all default houses, and have your first house from 1950, 2nd in 1960, etc
19:55:08 <supermop> you can still start a game in 1500, just all the houses will be the 1950 one (theater 'bug')
19:58:25 <Wolf01> I think it crashes or do weird things, cities expect a ROADTYPE_ROAD:SUBTYPE_NORMAL which is 0:0
19:58:51 <Samu> holy crap, trAIns is upgrading from monorail to maglev, it's that time again, but he got 500 trains, crazy
19:59:08 <Samu> upgrading one at a time is kinda slow
20:01:13 <Samu> down to 6 trains, what a crazy amount if irresponsibility
20:09:47 <supermop> Wolf01 i guess its a matter of making it more flexible in its expectations?
20:10:05 <Wolf01> Yes, but I lack of ideas
20:10:29 <supermop> or make it possible to redefine but not disable ROAD
20:10:56 <supermop> it seems like road types could have a property "is town road"
20:10:58 <Wolf01> The problem to redefine road is that you can make it incompatible with road
20:11:14 <supermop> if =1 town builds it
20:11:56 <supermop> if there are 16 that =1 maybe town still just builds the first, but any others so built will still get houses etc
20:12:47 <Wolf01> Yes, that could be an idea, when a town is found it choses randomly between those =1 and builds that one for the rest of the game
20:13:46 <Eddi|zuHause> i would propose that a town reruns that check regularly, and can pick a new roadtype to build once more get available
20:13:55 <supermop> additionally the townroad value could be not binary, but 0-7 or 0-255
20:14:23 <Eddi|zuHause> yeah, an automatic ranking, or grf author can provide one
20:14:27 <supermop> and 0 = town doesn't building
20:14:57 <supermop> else, town builds either the lowest (or maybe highest) value available
20:15:29 <Eddi|zuHause> town growth algorithm might want to replace existing roads with a new type
20:15:59 <supermop> if two are equal it either picks at random or based on order in which defined
20:16:15 <Eddi|zuHause> it should be deterministic
20:17:04 <supermop> author doesn't provide a value it gets 1 or whatever automatically
20:17:28 <supermop> or gets the value of its place order in definition
20:17:35 *** Rybeast has joined #openttd
20:18:33 <Rybeast> Hello! Sorry to be here again. Game failed to save - compressor initialisation failure - ideas?
20:20:26 <Rybeast> It didn't say. JUst said game failed to save - compressor initialisation failure
20:21:17 <Samu> hmm, does it have write access? :(
20:21:36 <Samu> gonna check what causes that error
20:21:39 <Rybeast> it's the only time it's failed to save at all :(
20:22:18 <Rybeast> I've just force quit the game and re-opened, but it says the save file has an error? 'Lilbizma returned error code'
20:22:37 <Borg> % git grep -E '.*init.*failure.*'
20:23:02 <Borg> I dont see such an error
20:23:19 <andythenorth> how about authors just don't disable ROAD?
20:23:23 <andythenorth> unless it's intended
20:23:27 <andythenorth> no guards, no checks
20:23:32 <andythenorth> I can 'break' all industries if I want
20:23:37 <andythenorth> nothing in spec stops m
20:24:12 <Rybeast> It ran until 2040 if that is worth anything?
20:24:24 <Samu> okay it failed to initialize lzma, that's something
20:25:52 <Samu> if (lzma_easy_encoder(&this->lzma, compression_level, LZMA_CHECK_CRC32) != LZMA_OK) SlError(STR_GAME_SAVELOAD_ERROR_BROKEN_INTERNAL_ERROR, "cannot initialize compressor");
20:26:31 <frosch123> if no ROAD, towns cannot build road, game does not start
20:26:36 <frosch123> same as when no houses are available
20:26:40 <Rybeast> What would have caused it to fail though? AS I say, it's been working the whole way
20:27:22 <supermop> frosch123 that works, but maybe its more fun to have alternatives to road?
20:27:25 <Samu> seems that during compression, something got corrupted?
20:27:40 <Samu> du u have ram issues or ?
20:28:20 <frosch123> supermop: a different appearance is still ROAD
20:29:36 <Samu> I wish which error it got :(
20:29:52 <andythenorth> loading a grf without ROAD is a bug
20:30:04 <andythenorth> bug / user error /s
20:30:50 <Samu> meh, all i know is that it was not the LZMA_OK, it was one of those 3
20:31:37 <Samu> - LZMA_MEM_ERROR: Memory allocation failed. I suspect it was this
20:32:03 <Samu> run 64-bit of openttd, etc. lol
20:32:18 <Rybeast> it's a brand new MacBook
20:32:51 <Samu> oh, i don't have a mac, i guess it could be something else
20:33:56 <Rybeast> :( Got to 2040 and built some amazing lines :/
20:35:28 <andythenorth> supermop: CC is a good idea
20:36:03 <Samu> LZMA_OK: Initialization succeeded. Use lzma_code() to * encode your data.
20:36:11 <Samu> LZMA_MEM_ERROR: Memory allocation failed.
20:36:18 <Samu> LZMA_OPTIONS_ERROR: The given compression preset is not * supported by this build of liblzma.
20:36:38 <Samu> LZMA_UNSUPPORTED_CHECK: The given check type is not * supported by this liblzma build.
20:36:55 <Samu> also unlikely, it's using LZMA_CHECK_CRC32
20:36:57 <LordAro> Samu: say something useful instead of listing the enumeration values
20:37:15 <Samu> LZMA_PROG_ERROR: One or more of the parameters have values * that will never be valid. For example, strm == NULL.
20:37:50 <Samu> i doubt the stream was null
20:38:46 <Rybeast> it's fine I guess. No way I'm going to get the level back, is there?
20:39:03 <LordAro> autosave is usually enabled by default...
20:39:59 <Rybeast> I've looked at the autosave folder - nothing there exept for the same error
20:40:33 *** Stimrol has joined #openttd
20:40:35 <Samu> maybe lzma for mac is bad? bad library or so? I dunno
20:41:03 <Samu> i dont have a mac to test :( can't help
20:41:51 <Samu> can u open openttd.cfg file
20:41:59 <Samu> seee which savegame_format u got there
20:43:00 <Samu> [misc] savegame_format =
20:43:34 <supermop> andythenorth little stripes or signs in cc could easily and discretely tell you who owns the wires, and the way the cc is drawn could help show what type of wires they are
20:43:44 <LordAro> pretty sure you can't change that setting while the game is running
20:43:48 <andythenorth> occasionally LZMA stops working on the mac, in official binaries
20:44:28 <Eddi|zuHause> if only there were a mac maintainer that could look at that kind of issues
20:44:33 <andythenorth> and if you don't notice quick enough, all autosaves are corrupt
20:44:45 <Rybeast> one of the autosaves has it all back
20:44:49 <andythenorth> and screen drawing is borked, black squares everywhere
20:44:58 <andythenorth> doesn't happen in self-compiled builds, or at least not to me
20:45:39 <LordAro> Eddi|zuHause: yeah, someone with a mac, and familiar with the ottd source
20:45:41 <andythenorth> it happens to my kids games
20:46:03 <andythenorth> and peter had one but it broke
20:52:37 <Samu> trAIns is still upgrading rails
20:53:16 <EarthlingKira> @Samu so, do you think that breakdowns are fine for high reliability ratings but should happen more often for low reliability ratings?
20:54:11 <Samu> hmm, didn't really test the first part
20:55:12 <Samu> gonna load some old saves I got here
20:55:15 *** sla_ro|master has joined #openttd
20:55:36 <Samu> and compare with 1 year of 1.7.2 for the same savegame
20:56:32 <Samu> ugh, i need to build again, this is gonna take a while again
21:04:25 <Samu> oho, it built already, this was faster now
21:12:48 <Samu> i think i need to give it more time, 1 year isn't enough
21:19:56 <Samu> top is with your patch, bottom is with 1.7.2
21:20:08 <Samu> reduced breakdowns on both
21:20:36 <Samu> it's almost unnnoticeable
21:21:45 <Samu> the impact of your patch seems like none
21:21:56 <Samu> breakdowns were off before 2051
21:22:27 <Samu> turned them on in 1st jan 2051, tons of aircraft went for autorenew, they were old
21:23:13 <Samu> no breakdowns vs reduced breakdowns, and the graph seems about equal to what it was before 2051
21:23:39 <Samu> now gonna test with normal breakdowns, brb
21:39:58 <Samu> for sake of completeness, gonna test without breakdowns :p
21:54:53 <Samu> first 2, reduced breakdowns, patch vs 1.7.2
21:55:02 <Samu> middle 2, normal breakdowns, patch vs 1.7.2
21:55:12 <Samu> last 2, no breakdowns, patch vs 1.7.2
21:57:45 <supermop> who wants to add stations to nml so I can update mlss?
22:00:12 <Samu> may have to test other vehicle types, too
23:04:27 <EarthlingKira> Thanks for the data and testings!
23:04:52 <EarthlingKira> You only compare the profit, right?
23:05:14 <EarthlingKira> I think breakdowns are most noticable when looking at profits when you do have really busy train tracks
23:05:37 <EarthlingKira> Otherwise their impact on profit is low (with and without my patch)
23:07:30 <EarthlingKira> I'd guess that AI doesn't build routes which are nearly as busy as routes built by humans
23:09:42 *** Smedles has joined #openttd
23:12:10 *** Progman has joined #openttd
23:27:29 <Samu> i wanted to try put this data in excel to create a graph
23:28:41 <Samu> a reliability of 65535 is the same as 100%
23:29:08 <Samu> every day CheckBreakdown is called, and 80 reliability is lost
23:30:45 <Samu> road vehicles service interval is a default of 150 days
23:33:13 <Samu> return (((uint16)r * b + b / 2) >> 16) < a;
23:33:30 <Samu> static inline bool Chance16I(const uint a, const uint b, const uint32 r)
23:38:33 <Samu> @calc (0xffff * 8 + 8 / 2) / (2**16)
23:41:49 <DorpsGek> Samu: 6.103515625e-05
23:46:13 *** Biolunar has joined #openttd
23:46:45 *** FLHerne has joined #openttd
23:52:08 <DorpsGek> Samu: 0.124986648356
23:59:02 <Samu> once every 8 times, add 4 to chance
23:59:11 <Samu> 8 times in this case, 8 days
continue to next day ⏵