IRC logs for #openttd on OFTC at 2019-05-12
⏴ go to previous day
00:39:21 *** Progman has joined #openttd
01:47:18 *** Eddi|zuHause has joined #openttd
02:09:42 *** Supercheese has joined #openttd
02:48:52 *** Wormnest has joined #openttd
03:25:47 *** ad5twoknebor[m] has quit IRC
03:25:53 *** hylshols7qui[m] has quit IRC
03:25:57 *** ist5shreawf[m] has quit IRC
03:25:58 *** yur3shmukcik[m] has quit IRC
04:13:57 *** ad5twoknebor[m] has joined #openttd
04:37:39 *** ist5shreawf[m] has joined #openttd
04:48:36 *** remmalj[m] has joined #openttd
04:57:24 *** fiddeldibu[m] has joined #openttd
05:08:46 *** ookfof[m] has joined #openttd
05:14:27 *** grag[m] has joined #openttd
05:21:14 *** godva[m] has joined #openttd
05:35:03 *** cawal[m] has joined #openttd
05:47:47 *** VoId23[m] has joined #openttd
05:55:40 *** kiwitree has joined #openttd
06:03:58 *** patrickm[m] has joined #openttd
06:05:58 <DorpsGek_II> [OpenTTD/OpenTTD] kiwitreekor commented on pull request #7575: Feature: Add industry production graph https://git.io/fjWTS
06:39:55 *** Supercheese has joined #openttd
06:55:29 *** yoltid[m] has joined #openttd
07:18:25 *** hylshols7qui[m] has joined #openttd
07:28:48 *** glothit7ok[m] has joined #openttd
07:49:44 *** yur3shmukcik[m] has joined #openttd
07:54:46 *** lapav[m] has joined #openttd
07:56:54 *** twom[m] has joined #openttd
08:00:52 *** olmvnec[m] has joined #openttd
08:17:16 *** pothyurf[m] has joined #openttd
08:22:33 *** andythenorth has joined #openttd
08:24:43 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN commented on pull request #7575: Feature: Add industry production graph https://git.io/fjWIC
08:31:00 *** tops[m] has joined #openttd
08:35:15 <andythenorth> oops, I was playing 6811
08:35:22 * andythenorth updates to master and compiles
08:37:22 <andythenorth> should I get my own server?
08:37:26 <andythenorth> infosec disaster :P
08:37:53 <andythenorth> oh the map selection buttons are broken in master
08:41:12 <andythenorth> padding is wrong
08:41:18 <andythenorth> I'll file an issue later
08:41:30 <nielsm> file off the excess padding
08:41:39 * andythenorth should go back to using stable releases :P
08:41:58 <andythenorth> using master is an easy way to get new admin obligations :|
08:46:39 *** iarp[m] has joined #openttd
09:01:40 *** labs[m] has joined #openttd
09:02:12 *** sla_ro|master has joined #openttd
09:04:21 *** nartir[m] has joined #openttd
09:10:24 <peter1138> Not sure anyone has touched the padding, but okay.
09:11:58 <peter1138> Right, bicycle tiem.
09:12:08 <nielsm> my patch for GUI scaling did some things with padding
09:12:13 <nielsm> but not explicitly in any window
09:17:13 <andythenorth> oops forgot to use 16 threads for make
09:17:20 * andythenorth waits for slow build
09:19:02 <LordAro> andythenorth: bro do you even sunk cost?
09:31:14 *** khavik[m] has joined #openttd
09:31:17 *** Progman has joined #openttd
09:51:06 <andythenorth> yeah padding is weird all over the place :)
09:52:00 <andythenorth> hmm some of it is not new
09:52:04 <andythenorth> present in 1.8.0
09:54:45 <Eddi|zuHause> andythenorth: alias make='make -j16'
09:55:14 <Eddi|zuHause> (i put something like that in a .profile somewhere)
09:55:33 <andythenorth> wonder if I can pass the -j flag through to python
09:55:50 <andythenorth> I actually need it to start a python MP worker pool
09:56:02 <andythenorth> I have my own stupid named arg for that
09:56:24 <Eddi|zuHause> put it in an environment variable?
09:56:48 <andythenorth> sounds plausible
09:57:58 <andythenorth> I should screenshot, but the vehicle window viewport has gained padding
09:58:05 <andythenorth> not broken, just different
09:58:27 <nielsm> def. do, and ideally mark it up
10:25:31 <V453000> yay one coun_veh_id() per engine
10:25:44 <V453000> not done coing yet though
10:29:30 *** jact[m] has joined #openttd
10:34:10 <Eddi|zuHause> "Structures in the center of every planet and moons now supply free oxygen" ... that should help a few things, i guess :p
10:35:05 <Wolf01> Regarding NRT, town roads get replaced with time or only new roads will be of the new buildable type?
10:36:42 <Wolf01> And if is the first one, does "road reconstruction" speed up the effect?
10:38:42 <andythenorth> Horse compile is now so slow
10:38:47 <nielsm> actually yeah the road reconstruction would make sense to upgrade the affected roads to the latest
10:39:01 <Eddi|zuHause> Wolf01: i would hope so, but i wouldn't expect it...
10:39:17 <andythenorth> so slow that I can open the compiled docs in a browser, view the sprites, and decide whether to ctrl-c the compile
10:39:48 <andythenorth> I think I've pushed nml about as far as it can realistically go
10:40:31 <Eddi|zuHause> andythenorth: have you implemented not-overwriting the sprite files yet, if they haven't changed?
10:40:44 <andythenorth> that looks worth doing
10:41:04 <andythenorth> after that, I need to decide whether to reimplement everything with layers
10:41:09 <Eddi|zuHause> just write to .png.new (or .new.png), do a cmp, and then overwrite
10:41:26 <andythenorth> peter said I shouldn't optimise with layers because it spanks openttd perfomance
10:41:40 <andythenorth> but it would cut down the number of sprites to encode substantially
10:41:46 <andythenorth> at the expense of massive complexity
10:42:03 <Eddi|zuHause> andythenorth: that sounds like the wrong tradeoff
10:42:38 <Eddi|zuHause> "i want to waste more execution time because i get bored during compile"???
10:42:46 <andythenorth> well I don't play much :P
10:43:01 <andythenorth> is there a preferred method for diffing bitmap files?
10:43:59 <andythenorth> 1m-2m is the worst cycle time
10:44:04 <andythenorth> it's not enough to make coffee
10:44:10 <andythenorth> it's long enough to be boring
10:44:14 <Eddi|zuHause> andythenorth: cmp will give you "same" or "different" on binary level
10:44:27 <Eddi|zuHause> andythenorth: you don't need anything more complex
10:44:48 <andythenorth> all the file handling is in python, so I need to use popen or something?
10:44:57 <andythenorth> or possibly python has a handler to cmp
10:45:05 <Eddi|zuHause> python might have a function for that
10:45:11 *** blim[m] has joined #openttd
10:46:19 <Eddi|zuHause> that sounds useful
10:46:22 <andythenorth> filecmp.cmpfiles looks like it should work
10:46:44 <andythenorth> so I replace all that don't match
10:47:07 <Eddi|zuHause> i would probably use cmp, at the end of the job that creates each file
10:49:14 <Eddi|zuHause> say: generate("{}.new.png".format(file)); if (not cmp("{}.png".format(file), "{}.new.png".format(file)) move("{}.new.png".format(file), "{}.png".format(file))
10:49:44 <Eddi|zuHause> that way you don't need to collect the list of all png files somewhere
10:49:54 <Eddi|zuHause> to then batch-compare them
10:50:07 <Eddi|zuHause> which would be a single-threaded bottleneck
10:50:30 <andythenorth> bit more intricate but eh ok
10:53:26 *** ciet[m] has joined #openttd
10:56:44 *** buggeas40d[m] has joined #openttd
10:59:45 *** andythenorth is now known as Guest2364
10:59:45 *** andythenorth has joined #openttd
10:59:51 *** cesar[m] has joined #openttd
11:08:11 <Wolf01> andythenorth: is NRT hog in bananas?
11:09:55 <andythenorth> only forums thread
11:27:38 <nielsm> hmm, the "hidden" flag on roadtypes doesn't prevent towns from building that type, does it?
11:27:55 <nielsm> so you could have a road type only buildable by towns and a different one only buildable by players
11:48:39 *** chomwitt has joined #openttd
11:49:56 *** HerzogDeXtEr has joined #openttd
11:53:41 *** gelignite has joined #openttd
12:26:02 <Eddi|zuHause> nielsm: that sounds weird at first, but it probably has some use...
12:32:17 *** Progman has joined #openttd
13:46:57 *** drac_boy has joined #openttd
14:47:15 *** andythenorth has joined #openttd
14:48:01 <andythenorth> is it lunchtime?
14:56:15 <LordAro> peter1138: you beat me
15:10:51 <Eddi|zuHause> LordAro: you probably deserved it.
15:16:49 <DorpsGek_II> [OpenTTD/website] auge8472 commented on pull request #85: Change: HTML5-elements and corresponding CSS-rules for the blog article template https://git.io/fjWYG
15:20:35 <andythenorth> now we watch me fail at make :P
15:22:07 *** frosch123 has joined #openttd
15:26:55 <peter1138> Oh are we combining mileage?
15:28:52 * andythenorth attempts to stop destroying nml spritecache
15:32:06 *** gareppa has joined #openttd
15:42:07 <LordAro> peter1138: if you like, we can say that you beat me yesterday, and i beat you today :p
15:44:39 <andythenorth> ok so not destroying the nml cache works
15:44:54 <andythenorth> if just one spritesheet changed
15:45:33 <andythenorth> shame it can't cache the parse :P
15:45:58 <Eddi|zuHause> you need partial compiles for that
15:47:15 <andythenorth> hmm these timings are a bit distorted, I deleted a bunch of trains
15:47:55 <andythenorth> comparing recent total run times is misleading, but the spritecache benefit is isolated and clear
15:48:07 <andythenorth> back down to 45s total
15:48:11 <andythenorth> I should delete more stuff
15:48:31 <andythenorth> should I split Iron Horse into separare grfs?
15:48:35 <andythenorth> Iron Horse - Engines
15:48:38 <andythenorth> Iron Horse - Wagons
15:48:43 <andythenorth> Iron Horse - Carriages
15:50:06 * andythenorth considers compiling 300 grfs
15:50:19 <andythenorth> the overhead probably eats the time saving
15:50:25 <andythenorth> but I have 16 thread units
15:50:49 <peter1138> LordAro, and I had a puncture today! ;(
15:50:58 <andythenorth> what is a puncture please?
15:51:36 * andythenorth has not had a puncture since about 2007
15:51:50 <andythenorth> I have ripped valves off the tube twice
15:52:13 <andythenorth> oh and I get a pinch puncture when the tube got between the rim and tyre wall
16:07:43 <Eddi|zuHause> andythenorth: i've rolled back from compiling each engine as 1 grf
16:07:51 *** gareppa has joined #openttd
16:07:53 <Eddi|zuHause> andythenorth: because it's just too much overhead
16:08:59 <Eddi|zuHause> aren't modern tyres self-repairing?
16:13:21 <andythenorth> mine are kevlared, never seen anything go through them
16:15:14 <LordAro> Eddi|zuHause: sometimes, depends on the tyre. there's usually some disadvantage to them
16:16:04 *** supermop_work has joined #openttd
16:18:02 <andythenorth> supermop_work: talk to me about British Rail hopper wagons 1960-2020
16:18:20 <supermop_work> this is some kind of fetish thing huh
16:18:57 <supermop_work> well my memory of seeing them only applies to 1997-1999
16:19:19 <supermop_work> but there seemed to be a variety of them on the rails around that time
16:19:44 <supermop_work> subsequent trips to the UK haven't afforded much freight stock viewing
16:20:10 <supermop_work> but i'd assume they'd still look stubbier than american ones?
16:25:03 <andythenorth> if these were 2 different types in game, what would they be called?
16:31:56 <supermop_work> they both empty through chutes on the bottom?
16:35:10 <andythenorth> one has auto-unloading, one doesn't
16:35:14 <peter1138> So I wondered about roadtype/town callback stuff.
16:35:18 <peter1138> But it's a bit non-obvious.
16:35:33 <peter1138> Like, who gets to decide?
16:36:18 <peter1138> Shall I buy a sous vide machine? Seems to be some kind of fad.
16:37:51 <andythenorth> it's nice to have a fa
16:37:56 *** Wormnest has joined #openttd
16:38:45 <peter1138> And why would one need to be Alexa-compatible? o_O
16:39:20 <andythenorth> so Amazone know what you're eating
17:18:07 *** snail_UES_ has joined #openttd
17:18:48 <Eddi|zuHause> damn you stalebot! *throws fist*
17:31:23 <peter1138> Well it hadn't been updated for 3 months...
17:51:53 <dwfreed> peter1138: sous vide is a common cooking method in high end restaurants; it's become popular lately because people want the high end restaurant experience, but without the cost of going out to eat
17:51:56 *** kiwitree has joined #openttd
18:42:06 <peter1138> I know what it is :-)
18:54:10 <snail_UES_> as I was playtesting my set, I noticed lots of settings concerning the economy are so unrealistic
18:54:33 <snail_UES_> inflation raises costs more than profits… infrastructure costs are more than proportional to the scale of your network
18:54:43 <snail_UES_> this is totally NOT how the real world works
19:01:25 <peter1138> inflation raises costs so that the game gets harder.
19:03:07 *** andythenorth has joined #openttd
19:04:17 <snail_UES_> so in order to make the game get harder, we implement something inaccurate?
19:04:24 <Eddi|zuHause> snail_UES_: infrastructure costs are also meant to simulate your overhead. if you employ 2 workers, you must pay double the wage, but if you employ 20 workers, you must also employ a manager that earns a multiple of a worker wage
19:04:31 <snail_UES_> wouldn’t it be better to place something like that in the difficulty levels or so?
19:04:38 * andythenorth sat in the hot tub for 1 hour at 40º
19:04:47 <andythenorth> now I'm spaced out :D
19:05:06 <snail_UES_> Eddi|zuHause: yes, but 20 workers will be able to do a better job than 2 workers * 10
19:05:12 <snail_UES_> ever heard of economies of scale? :P
19:05:27 <snail_UES_> if what you said was correct, there wouldn’t be any largecap companies in the world
19:05:36 <Eddi|zuHause> snail_UES_: and the inflation difference between income and costs must be dealt with by improved technology
19:06:14 <snail_UES_> Eddi|zuHause: inflation raises prices first, wages later. So what’s implemented in the game is totally unrealistic
19:06:31 <snail_UES_> it’s like wages were raised before the prices of the goods…
19:06:37 <Eddi|zuHause> snail_UES_: you get faster vehicles
19:06:39 <snail_UES_> that’s just not how it works
19:06:59 <snail_UES_> fatser vehicles will have a higher maintenance cost, so that’s already taken into account
19:07:02 <Eddi|zuHause> snail_UES_: it's not a perfect simulation
19:07:08 <snail_UES_> and (usually) higher fuel comsumption
19:07:16 <snail_UES_> I know it’s not perfect
19:07:16 <Eddi|zuHause> snail_UES_: it's a fairly simple formula
19:07:20 <snail_UES_> but this is blatantly broken
19:07:36 <Eddi|zuHause> snail_UES_: again, you're meant to balance this out by better technology
19:07:38 <snail_UES_> I mean it’s like introducing concepts that are broken to begin with...
19:07:44 <_dp_> it's not an economy simulator to be realistic :p
19:08:03 <Eddi|zuHause> no economy simulator is realistic :)
19:08:22 * andythenorth wonders about reading logs :P
19:08:26 <snail_UES_> but at least shouldn’t it “try its best” to be realistic?
19:08:28 <andythenorth> is something else broken?
19:08:35 <Wolf01> Also in 2000 you start with the same money you had in 1950, which allow you to buy like 5 times less things
19:08:36 <snail_UES_> here it’s just doing something totally weird...
19:08:55 <andythenorth> snail_UES_: IMHO, no. It should do its best to offer content APIs to authors of mods
19:08:56 <Eddi|zuHause> snail_UES_: well, you're welcome to come up with a better system?
19:09:06 <nielsm> the inflation is just a mechanism to change the ratio between costs and incomes without having to resort to fractional currency
19:09:07 <snail_UES_> Wolf01: correct, the initial loan should be balanced according to the costs of the period you start at
19:09:11 <andythenorth> is inflation a thing?
19:09:18 <andythenorth> I thought it was deprecated?
19:09:25 * andythenorth is confused, yet again
19:09:28 <Wolf01> Joining a game with inflation enabled after 50 or even 100 years is literally unplayable
19:09:32 <andythenorth> I'll get my coat
19:09:38 <nielsm> the inflation setting is not deprecated, afaik
19:09:43 <snail_UES_> nielsm: ok, but in reality, prices are what inflation raises first
19:09:54 <snail_UES_> so it should benefit a transportation company
19:09:57 <andythenorth> yes, but we don't fix bug reports about inflation
19:10:00 <nielsm> snail_UES_ yes but that's irrelevant for this
19:10:03 <andythenorth> we tell the player to turn inflation off
19:10:06 <snail_UES_> wages are raised later as a result of this
19:10:41 <andythenorth> inflation would make debugging newgrfs really horrible, please don't encourage it :P
19:10:50 <nielsm> it's a game mechanic intended to introduce difficulty, it's not a game mechanic intended to simulate a real world mechanism
19:11:01 <Eddi|zuHause> snail_UES_: that makes no difference, whether the chicken or the egg was first
19:11:18 <snail_UES_> nielsm: ok perfect. But then let’s not call it “inflation”… let’s call it, “difficulty settings” or something like that
19:11:33 <_dp_> wtf is a difficulty in a sandbox?
19:11:41 <andythenorth> "broken: on | off"
19:12:02 <andythenorth> I need more emojis
19:12:13 <Eddi|zuHause> i think we can safely ignore all lines from andythenorth that include the word "broken"
19:12:22 <andythenorth> is there an emoji for "I am just talking crap to see how it sounds" ?
19:12:49 <snail_UES_> I’ll add two notes to my set “donotreadme”… please turn inflation off, and please turn infrastructure costs off
19:12:50 <andythenorth> it's so hard to disambiguate from "no I really mean this and I am using exactly the same tone in irc"
19:13:01 <nielsm> the problem with inflation as implemented is that it bases the rate on the loan interest rate in a way so the lower interest rates makes the game harder later on, while the higher interest rates makes the game not quite as much harder later on
19:13:10 <andythenorth> I think TB has that problem too ^^
19:13:26 <nielsm> since the inflation rates are interest% and (interest+1)%
19:13:41 <nielsm> so on a higher interest rate the ratio between the two numbers is lower
19:13:42 <andythenorth> isn't this a GS thing?
19:13:54 <_dp_> somehow if you add a goal every thing that's said to make game more difficult end up just being stupid instead
19:14:20 <snail_UES_> nielsm: what are these two quantities ( interest% and (interest+1)% ) ?
19:14:37 <snail_UES_> are they used for revenues and costs, respectively?
19:14:39 <nielsm> snail_UES_: the rate at which incomes rises and the rate at which costs rise
19:14:46 <snail_UES_> ah, so it’s totally broken :)
19:14:53 <andythenorth> why is inflation desired?
19:14:56 <snail_UES_> it should be: interest% and (interest * 1.1)%
19:15:01 <nielsm> it's called "inflation" because it makes numbers go up over time
19:15:14 <nielsm> yeah but that would make the calculations much harder
19:15:17 <snail_UES_> inflation is a nice realistic thing, but it doesn’t work the way it’s implemented
19:15:29 <snail_UES_> nielsm: so we should use the same number on both sides...
19:15:46 <snail_UES_> or at least have a flag or a setting to play on the difference
19:15:55 <nielsm> if the revenue and cost rates are the same then it's just pointless to enable
19:16:06 <nielsm> apart from savings becoming less worth over time
19:16:26 <nielsm> (effectively negative interest on savings)
19:16:33 <andythenorth> is it just a multiplier to all currency displays?
19:16:39 <nielsm> (well, liquid capital?)
19:16:55 <snail_UES_> no, it would still have a meaning. Your interest on your loan would be more and more irrelevant with high inflation
19:16:57 <andythenorth> IT WILL NEVER WORK WITH DAYLENGTH
19:17:11 <nielsm> no, inflation is a multiplier to actual revenue and cost figures
19:17:12 <snail_UES_> no inflation = your interest on your initial loan will be a huge burden
19:17:24 <snail_UES_> lots of inflation = after a few years it becomes irrelevant
19:17:44 <_dp_> snail_UES_, wut? interest is irrelevant no matter how high it is
19:18:03 <andythenorth> just having an 8/8 wagon won't work
19:18:03 <snail_UES_> this would mean that your interest will be lower and lower as a percentage of your income
19:18:10 <nielsm> but yes the max loan (and maybe loan take/repay amounts) should be scaled up by inflation to make sense
19:18:40 <snail_UES_> and the interest you pay on outstanding loan will of course be untouched by inflation (we can assume fixed rates)
19:18:58 <snail_UES_> so low inflation —> interest is always a big factor, “go easy on loans"
19:19:28 <snail_UES_> high inflation —> interest will fade away as a percentage of your revenues as time goes on, therefore you can be a bit more aggressive when borrowing
19:20:05 <snail_UES_> I think something like this would make sense. And also allow the player to choose the differential between revenues and cost inflation
19:20:20 <_dp_> with how much money you can make interest needs to be like 200% or smth to make any difference at all
19:20:33 <snail_UES_> _dp_: not in the very early years
19:20:49 <andythenorth> ok maybe a 6/8 version then
19:20:53 <snail_UES_> when you start, you can be a bit more aggressive if inflation is high, you can’t if it’s low or non-existent
19:21:17 <snail_UES_> but I agree, later in the game it’d make a smaller difference (your revenues would be large enough)
19:21:33 <_dp_> snail_UES_, early years aren't that different on cost/profit
19:21:59 <snail_UES_> but when you have only 1 or 2 lines, interest expense can be quite significant to you
19:23:40 <_dp_> snail_UES_, if your goal is money first train about pays for itself on a first run
19:23:53 <_dp_> snail_UES_, I never even bother to repay loans
19:24:14 * andythenorth wonders about alternative mechanics
19:24:23 <andythenorth> like, the game could dump random costs on you
19:24:29 <andythenorth> 'GDPR compliance'
19:24:44 <andythenorth> 'Euro 6 engines'
19:24:44 <_dp_> andythenorth, ZOMBIES! xD
19:24:46 <snail_UES_> andythenorth: there could be taxes as well
19:24:56 <andythenorth> there's no dividend flow either
19:36:22 <peter1138> 18:18 < nielsm> but yes the max loan (and maybe loan take/repay amounts) should be scaled up by inflation to make sense
19:38:40 <snail_UES_> peter1138: but interest payments on outstanding loan shouldn’t (if we assume fixed rates)
19:38:53 <snail_UES_> also, the design mistake isn’t there. It’s that costs increase more than revenues
19:41:23 <V453000> WTF :D how could os.makedirs(output_folder, exist_ok = True) give me TypeError: makedirs() got an unexpected keyword argument 'exist_ok' ... it works in my other scripts ._. the path seems right ('Output folder is:', 'D:\\OpenTTD_repositories\\NUTS-git\\scripts\\..\\src-generated')
19:41:42 <V453000> the ".." does not seem to be a problem
19:43:53 <V453000> D:\OpenTTD_repositories\NUTS-git\scripts>python --version this gives me
19:43:56 <V453000> so that should be right
19:44:02 <V453000> the print() would not work otherwise
19:44:25 <dwfreed> "Note makedirs() will become confused if the path elements to create include pardir (eg. “..” on UNIX systems)."
19:45:42 <dwfreed> since you have python3.6, you should use a Path instead of a string, and .. will be handled for you
19:46:01 <snail_UES_> another design mistake about inflation… it’s always constant
19:46:18 <snail_UES_> this generates a compounding effect, so that prices in 2020 in a game started in 1880 will be ridiculously high
19:47:00 <nielsm> solution: let GS alter the rate!
19:47:32 <V453000> I wasn't aware a path type exists even :D
19:47:33 <snail_UES_> solution: start with a constant rate, and apply random changes (could be generated by a normal distribution) every year
19:47:44 <dwfreed> V453000: was newly added in 3.6
19:48:15 <nielsm> peter1138: maybe the max loan is scaled then, but the starting cash amount and loan take/repay step isn't
19:49:04 <nielsm> (someone is going to say "what if your loan amount is not a multiple of the loan repay amount" and the answer is "just repay the rest if the repay amount is larger than current loan")
19:50:53 <Rubidium> peter1138: isn't max loan already inflation scaled (in I think 50k blocks)?
19:51:45 <andythenorth> hi Rubidium :D :o
19:52:30 <V453000> hm, os.path.normpath seemed correct but gives me the same thing
19:52:43 <milek7> 'New in version 3.2: The exist_ok parameter.'
19:53:02 <V453000> yeah and I'm on 3.6.5 so that should work?
19:53:21 <V453000> it always worked for me, I don't understand what is wrong :D
19:53:37 <V453000> I'll just remove it for now, it's just one folder and it already exists but wtf :d
19:53:49 <V453000> I guess I could check if the path of the folder exists manually instead
19:54:24 <andythenorth> if not os.path.exists(graphics_output_path):
19:54:24 <andythenorth> os.mkdir(graphics_output_path)
19:54:33 <andythenorth> is code I was using today
19:55:19 <Rubidium> snail_UES_: inflation corrected prices for airline tickets have halved between 1980 and 2010, so it's definitely real (for some forms of transport) that the ticket prices were raised less than inflation. Prices of aircraft were raised faster than inflation though (the 747's different models in this case)
19:55:19 <V453000> well, works. Strange, but fuck it for now
19:55:58 <snail_UES_> Rubidium: yes, this has a lot to do with external factors such as oil price
19:56:43 <snail_UES_> if we want to keep it simple, such as in a simulation, we should at least raise revenues and costs at the same rate, which would be correct “on average”. It wouldn’t make sense to privilege one vs. the other
19:58:10 <snail_UES_> Rubidium: as for prices of aircraft, it was true for the 747 but not for the smaller planes, where there was more and more competition (think Embraer)
19:58:30 <LordAro> Rubidium? in #openttd? what year is it?
19:58:49 <snail_UES_> also, lower oil prices affect not only ticket prices, but also fuel prices. This would be a lower cost for an airline company
19:59:42 <michi_cc> snail_UES_: Inflation in OTTD is in no way about realism. Incoming is growing anyway due to faster vehicles and bigger towns. OTTD inflation just buffers that grow a bit.
20:00:01 <LordAro> V453000: that error definitely implies it's not using your 3.6 python install - sure it's not using a 2.7 by accident? (`python` vs `python3`)
20:00:35 <snail_UES_> so let’s not call it inflation… let’s call it, “exploding costs to increase difficulty level”
20:00:56 <LordAro> V453000: adding `print(sys.path)` just before should tell you where it's looking for python files
20:01:05 <snail_UES_> in one of my tests (started in 1880), in year 1995, a mid-size narrow gauge railcar cost as much as about 500 new VW Golf of the same era…!!
20:01:31 <snail_UES_> this is like… beyond laughable
20:01:55 <Rubidium> are those from the same NewGRF?
20:02:16 <snail_UES_> the VW Golf is from a NewGRF called “real world” :)
20:02:42 <snail_UES_> and the railcar is from my own NewGRF, with its cost scaled so that it would match the real world in a game started in that year (1995)
20:03:17 <snail_UES_> the issue was that the railcar’s cost got compounded so much to become astronomical…
20:03:38 <michi_cc> But does the income from a trip with this railcar match real world as well? Otherwise, matching costs is a bit pointless.
20:03:43 <nielsm> aren't most newgrf vehicles also cost-scaled by (non-in-game) inflation already? i.e. if you compare the inflation-adjusted prices of real world locomotives in 1880 and 1995 you wouldn't see the same price increase you see in non-in-game inflation adjusted prices in ttd
20:05:40 <snail_UES_> michi_cc: maybe it’s matched in-game, but it’s the figure that it a bit silly. A railcar costing DM 13.8M in 1995? Makes me think of Weimar republic prices
20:05:57 <V453000> LordAro: yeah its looking in 2.7
20:06:02 <nielsm> one example, a danish steam loco from 1909 was purchased from factory for 64,000 kr, adjussted for inflation to 2018 prices that's about 4.2 million kr
20:06:06 <V453000> how did I mess this up :D
20:06:48 <nielsm> stats are 60 km/h, 44.8 tons weight, not sure how to determine a power value from the data I have here :P
20:07:01 <snail_UES_> nielsm: exactly. The prices of locomotives went down, because of better technology, which makes it easier to build them
20:07:27 <snail_UES_> nielsm: if you give me firebox surface, grate area, boiler pressure, and superheating surface, I can give you an estimate :)
20:08:29 <snail_UES_> the issue here is that, if I include a 1995 railcar in my trainset, I will set its price around its real-world figure (in 1995)
20:08:54 <nielsm> only have boiler pressure, 12.8 kg/cm^3
20:08:54 <snail_UES_> but if I start a game in 1880, the very same amount I set in my code gets compounded by inflation, therefore hitting an astronomical figure
20:09:08 <snail_UES_> ah… without grate area, it’s going to be tricky
20:09:37 <snail_UES_> I found the product of grate area * boiler pressure to be an acceptable way to estimate a steamer’s power (statistically speaking)
20:09:58 <snail_UES_> with firebox, tubes, and superheating surface to be useful to fine-tune the number
20:10:06 <michi_cc> Well, of course inflation goes bonkers if you apply it twice; once by yourself, once by the game.
20:10:26 <snail_UES_> michi_cc: I don’t apply it by myself. I’m only setting a price as of 1995
20:11:00 <snail_UES_> the way the game currently works, there is no way to have a resonable 1995 price always displayed, because it’d depend on the year a game starts
20:12:01 <snail_UES_> nielsm: cute steamer! but I can’t understand Danish… :(
20:12:12 <snail_UES_> I’ll try with google translate
20:12:21 <nielsm> and a factory new siemens vectron electric loco is estimated to be around 25-30 million kr
20:12:30 <michi_cc> snail_UES_: But you do. If the same vehicle would have been invented in 1985, it would have had a different price, wouldn't it?
20:13:06 <nielsm> so the cost per power and speed unit seems to have dropped a bunch
20:13:34 <snail_UES_> michi_cc: I would always give a “nominal” price that’s consistent with the introduction time’s price index
20:13:55 <snail_UES_> nielsm: yes, that’s my point!
20:14:29 <snail_UES_> so it’d actually make sense to raise income more than costs… because of what you just said (engines have become cheaper through time)
20:15:05 <snail_UES_> nielsm: looks like there’s no grate area for that cutie. Have you tried JDH Smith’s tables?
20:15:14 <_dp_> what's the point of realistic prices? are you making addon for the game or history museum?
20:15:39 <snail_UES_> no, but having a railcar costing like 10M euros? Seriously?
20:16:38 <snail_UES_> anyway one can live with that. The issue is costs growing more than revenues, making economies of scale impossible (this is especially true for infrastructure maintenance costs)
20:16:49 <andythenorth> why is the inflation switch still present?
20:16:54 <andythenorth> can I put a PR in to remove it?
20:16:56 <snail_UES_> I totally get the idea of making the game harder
20:17:08 <snail_UES_> I just think this is a bit weird way to achieve that...
20:17:52 <_dp_> openttd has shitton of stupid settings, just don't use then :p
20:18:05 <andythenorth> snail_UES_ found any narrow gauge cryo tankers in your research? o_O
20:18:29 <andythenorth> hmm stupid mac touchbar crashed again, let's restart it
20:18:37 <michi_cc> snail_UES_: Revenues are also growing because towns grow and produce more pax/mail and vehicles get faster, both lead to increasing revenue (independent of any inflation setting).
20:20:45 <michi_cc> In the game context you can't view game inflation independent from this "natural" grow. This mechanism is at odds with reality, but then OTTD is a game and not a reality simulator.
20:21:33 <snail_UES_> I found some threads speaking about this, I think it’s a problem. But wouldn’t it be better to attack it by trying to play with the mechanism that generates pax and mail, and grows towns?
20:21:46 <michi_cc> TL;DR: Some game rules are simply incompatible with a reality simulator, even if you try to simulate a very nice NewGRF reality :p
20:21:58 <snail_UES_> fixing it with inflation seems to me like… introducing two flaws and hope they can balance each other out
20:22:19 <andythenorth> it was an accident, not a design?
20:22:31 <andythenorth> inflation seems to be clown shoes early days stuff
20:22:43 <snail_UES_> andythenorth: I think so
20:22:45 <michi_cc> I think it even was in TTO.
20:23:26 <andythenorth> ok it's an accident
20:23:35 <andythenorth> no account for variable start dates
20:23:38 <andythenorth> no account for grf
20:23:59 <michi_cc> And they got the difficulty wrong even back then. Easy preset had 2% and hard 4%, but 4% is actually easier than 2%.
20:24:00 <andythenorth> presumably we're not allowed to delete it if it was in TTO?
20:24:10 <andythenorth> we have to keep ALL TTO/TTD features?
20:24:25 <michi_cc> andythenorth: Why do you think there is a setting?
20:24:33 <LordAro> default can be changed to "disabled"
20:24:49 <andythenorth> isn't it already?
20:25:56 <andythenorth> oh yeah, one of my kids was playing and it ruined his game
20:26:06 <andythenorth> it just makes the game totally unusable
20:26:20 <Wolf01> Still talking about inflation?
20:26:31 <andythenorth> much talk, little action :D
20:26:36 <LordAro> who's the one that's running the multihundred year game? they mentioned it as well
20:26:50 <andythenorth> yeah child #1 is addicted to ffwd
20:26:57 <andythenorth> so he gets through 500 years quickly
20:27:07 <Wolf01> I'm +1 for changing the default to disabled
20:27:16 <LordAro> andythenorth: blame peter1138 :p
20:27:23 <andythenorth> are we sure it's not disabled already?
20:27:29 <andythenorth> LordAro: seems unproductive :P
20:27:55 <snail_UES_> andythenorth: I think it’s enabled by default
20:28:09 <snail_UES_> I’d agree to change it to disabled
20:28:39 <andythenorth> I don't know who's in charge here :D
20:28:42 <andythenorth> [but it's not me]
20:29:04 *** Flipp3rrr has joined #openttd
20:30:11 <LordAro> water at map edges :p
20:30:39 <_dp_> btw, adding a warning for deprecated settings would be nice
20:31:07 <_dp_> helped someone recently who had order.selectgoods off
20:31:22 <nielsm> what would that warning consist of?
20:31:52 <nielsm> big red text in settings window?
20:31:56 <nielsm> popup when starting the game?
20:32:00 <LordAro> "this setting is only intended for compatibility with old savegames, use with caution"
20:32:18 <_dp_> nielsm, eh? just a warning when savegame has hidden settings in non-default values
20:32:21 <LordAro> extra tab/dropdown in the settings window could work
20:32:56 <nielsm> basic/advanced/expert/expert+deprecated
20:33:17 <nielsm> (except deprecated settings changed from default always show, and with a warning sign on)
20:33:36 <_dp_> nielsm, yeah, something like that
20:41:01 <_dp_> btw, isn't selectgoods an original ttd thing that was removed?
20:42:13 <nielsm> the original behaviour is that a station gets all goods from all producers immediately
20:42:14 <snail_UES_> it looks like the D(II) from jdh smith’s tables
20:42:22 <nielsm> selectgoods as a setting was a ttdpatch thing
20:43:29 *** Flipp3rr_ has joined #openttd
20:45:57 <snail_UES_> ok, your steamer should have a power of around 750HP…
20:46:08 <snail_UES_> if you want a confidence interval, I’d say between 700 and 800HP
20:46:22 <snail_UES_> tractive effort should be around 83kN
20:48:05 <nielsm> and so for the vectron comparison, the vectron can run 200 km/h with 8600 hp, so 3.3x speed and 11x power for 6x price
20:51:57 <nielsm> (and also requires only a driver, no fireman)
20:53:52 <andythenorth> also no-one to drop and clean the ashpan
20:53:56 <andythenorth> nor clean the smokebox
20:54:07 <snail_UES_> right… which makes it a bit weird to have costs increase more than revenues
20:56:37 <_dp_> nielsm, well, since it was done already inflation can just follow selectgoods example :p
21:00:28 <peter1138> 18:38 < snail_UES_> also, the design mistake isn’t there. It’s that costs increase more than revenues
21:00:48 <snail_UES_> peter1138: which is not in reality :)
21:00:59 <peter1138> ^ Without that, inflation would be just a multiplier with no other effect.
21:01:15 <andythenorth> it would just make the numbers arbitrarily different
21:01:19 <snail_UES_> ok, so it’s mislabeled
21:01:42 <snail_UES_> it could very well exist as a difficulty setting
21:01:53 <snail_UES_> “Allow costs to grow at a higher rate than revenues"
21:02:10 <snail_UES_> that’d be perfectly fine and the player would exactly know what he’s going for
21:02:40 <snail_UES_> just… it’s not what inflation actually does. So it’s mislabeled to the least
21:04:30 <Flipp3rr_> Hello, can anyone help me with IRC? I wanted to connect to another channel but got the error <irc.rizon.net> Error(404): #jb Cannot send to channel
21:05:50 <Flipp3rr_> Nevermind, it works now.
21:11:32 *** tokai|noir has joined #openttd
21:11:32 *** ChanServ sets mode: +v tokai|noir
21:12:06 <peter1138> LordAro, crashes in master too.
21:12:18 <peter1138> Oh, I'm in a PR. /me double checks
21:16:02 <peter1138> The PR I was in fixes on of the b0rks.
21:16:34 <peter1138> Eh, no it doesn't, that was merged. Hmm.
21:16:39 <peter1138> ANother k-d tree loading bug, then. :/
21:17:18 <peter1138> Which is a separate issue to loading in 1.9.x
21:17:27 <LordAro> 1.9 doesn't have- heh.
21:18:23 <peter1138> RebuildViewportKdtree() still being called before objects are converted in the map array.
21:18:41 <LordAro> andythenorth: not before August, imo
21:19:45 <peter1138> nielsm, c9fe6e7b8 needs to be applied to the other tile types.
21:22:03 <peter1138> Yeah, seems to fix the k-d bug. Now to the actual bug :p
21:23:47 <peter1138> Um. RailTileType is 2. Which isn't one of the enum values.
21:23:50 <LordAro> given one of these needs to be backported, i'd suggest 2 PRs
21:25:55 <peter1138> Sure, I wouldn't do the k-d tree fix together.
21:28:48 <peter1138> - RAIL_TILE_WAYPOINT = 2, ///< Waypoint (X or Y direction)
21:28:52 <peter1138> So, a long long time ago...
21:30:02 <peter1138> Trying to find where they are converted :-)
21:30:46 *** andythenorth has joined #openttd
21:30:49 <peter1138> void MoveWaypointsToBaseStations()
21:33:30 <peter1138> That already happens before. Hmm.
21:33:59 <supermop_work> andythenorth: requires game to adjust cost of labor vs cost of materials over time
21:35:19 <andythenorth> also needs some lolz
21:37:10 <LordAro> well obviously it's not doing its job properly
21:39:01 <peter1138> Ish. It's converting most of them.
21:39:31 <peter1138> So old waypoint 1 has an xy of 48731, but that is not a waypoint tile.
21:39:53 <peter1138> This tile is 50011, with index 1
21:41:11 *** Flipp3rrr has joined #openttd
21:43:29 *** Flipp3rrr has joined #openttd
21:48:50 <glx> [20:30:41] <_dp_> btw, adding a warning for deprecated settings would be nice [20:31:09] <_dp_> helped someone recently who had order.selectgoods off <-- not deprecated, just removed from the GUI (don't know why, but last time I searched a long time without finding it)
21:53:30 <peter1138> I don't know why either.
21:53:42 <peter1138> Isn't that what the different settings levels are for? :/
21:53:53 <peter1138> I keep meaning to make a PR to put everything back into the GUI.
21:55:45 <Eddi|zuHause> (just picking a random sentence from the backlog that i haven't fully read) <snail_UES_> no inflation = your interest on your initial loan will be a huge burden <-- this would be true, if we had any sensible kind of game balance... but income is so disconnected from costs that you have repaid your starting loan long before inflation is even an issue
21:57:20 <glx> yes I know that, but I don't remember what was the reason to remove it from GUI
21:57:38 <snail_UES_> depends on the set vehicles...
21:58:09 <snail_UES_> if you use vehicles with low capacities (such as NG trains), in the very few years, interest paid are a (small) factor
21:58:18 <peter1138> Yeah most of those should still be in the UI.
21:58:26 <_dp_> (svn r26528) -Remove: A bunch of archaic settings from the advanced s
21:58:26 <snail_UES_> but it’s true that after having 2 or 3 profitable lines, it becomes negligible
21:58:55 <snail_UES_> my point is, I’m not against this feature, but it should be called for what it is. A difficulty setting. Not an inflation setting
21:58:58 <glx> but the categorisation of settings in GUI is almost wrong anyway
21:59:02 <peter1138> snail_UES_, everyone making NewGRFs designs them to be played with inflation off.
21:59:17 <peter1138> snail_UES_, inflation IS a difficulty setting.
21:59:35 <peter1138> We don't have a separate section for difficulty settings.
21:59:37 <snail_UES_> peter1138: but “inflation” does not behave the way it’s in the game
21:59:41 <Eddi|zuHause> snail_UES_: there's more difficulty settings, you gotta have a name for it somehow...
21:59:49 <glx> and inflation was made for the original game time frame
21:59:57 <snail_UES_> if you call this “inflation”, then it’s a design mistake
22:00:01 <Eddi|zuHause> snail_UES_: lots of things don't behave like they are in the game, that's not an argument
22:00:08 <snail_UES_> if you call this “difficulty setting”, then it’s a mislabeling
22:00:18 <peter1138> snail_UES_, you realise this inflation was in the original TTD, right?
22:00:19 <_dp_> "difficulty" setting for noobs :p
22:00:29 <snail_UES_> peter1138: yes, I’m playing this thing since TTO
22:00:31 <peter1138> And it was called inflation back then.
22:00:43 <nielsm> was it even a setting originally?
22:00:51 <snail_UES_> that’s ok, so it means it’s been broken since the beginning. Not arguing with that
22:00:59 <Eddi|zuHause> it was bundled with the other difficulty settings
22:01:05 <snail_UES_> so let’s say that we inherited a design mistake
22:01:10 <Eddi|zuHause> like industry density
22:01:20 <glx> it's just worse now with custom start date and late ending
22:01:28 <snail_UES_> but since we now have OTTD, we have the chance to fix this… or at least name it correctly
22:01:51 <snail_UES_> glx: exactly. With inflation on, prices in year X will depend on when you started the game
22:02:04 <Eddi|zuHause> snail_UES_: there's lots of things i would fix about it, but the 1% difference between cost and income is probably very low on that list
22:02:27 <Eddi|zuHause> snail_UES_: the start date dependency is a problem
22:02:43 <snail_UES_> yep and it’s exacerbated by constant inflation
22:02:51 <snail_UES_> through the compounding effect
22:02:59 <Eddi|zuHause> nielsm: the interest rate is the inflation setting
22:02:59 <glx> would maybe ok if end date is always start date + 100 years
22:03:01 <nielsm> inflation is not a difficulty setting (or any other kind of setting)
22:03:06 <nielsm> Eddi|zuHause yes but you can't disable it
22:03:09 <snail_UES_> if inflation were a random variable that changes once a year, with average of (say) 2%, it would be less of a problem
22:03:15 <nielsm> ttdpatch let you disable inflation
22:03:23 <michi_cc> glx: Isn't inflation duration already limited?
22:03:41 <Eddi|zuHause> nielsm: and technically 2% is more difficult than 4%
22:04:05 <nielsm> but 2% is the "easy" and 4% the "hard"
22:04:46 <snail_UES_> in reality, prices don’t explode because we have years of deflation (or negative inflation)
22:04:57 <michi_cc> It might make sense to define inflation to always start in e.g. 1920. If you start earlier, inflation start will still be that year, if you start later, the game will pre-inflate the money till the start date.
22:05:13 <snail_UES_> if inflation were a random variable in the game, the effects would be more plausible
22:05:57 <Eddi|zuHause> michi_cc: yeah, that would be a start
22:06:32 <Eddi|zuHause> michi_cc: and maybe stop it earlier, like 2050 instead of 2090 (170 years)
22:06:52 <Eddi|zuHause> michi_cc: or whenever the vehicle progression gets frozen
22:16:41 <peter1138> michi_cc, yes, it is limited to 170 years, just the start date isn't fixed.
22:17:01 <peter1138> Hmm, managed to make that savegame load by doing the waypoint conversion "in reverse">
22:17:11 <peter1138> (Map array loop instead of old waypoint loop)
22:18:38 <andythenorth> delete everything :P
22:18:52 <andythenorth> I am confused now about the right way to newgrf :P
22:19:00 <andythenorth> should I be using layers, or composited sprites?
22:19:08 <andythenorth> (pre-composited)
22:19:31 <nielsm> one is faster compiles, other is better in-game performance?
22:20:22 <Eddi|zuHause> if it's a compile-time to run-time tradeoff, run-time should win every time. but even in a memory to run-time tradeoff, the run-time is the heavier one usually
22:25:26 <peter1138> Apparently the sprite cache can "explode" without layers
22:26:23 <peter1138> And now this save asserts in OrderList::DebugCheckSanity
22:26:32 <peter1138> assert(this->timetable_duration == check_timetable_duration);
22:39:59 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN opened pull request #7588: Fix e8d397e4ee: Avoid using RemapCoords2 during savegame conversion. https://git.io/fjWnx
22:44:44 <LordAro> why is RemapCoords2 so dangerous? what advantage does it have?
22:45:19 <nielsm> it checks for foundations
22:45:19 <Eddi|zuHause> lots of preconditions might be violated during savegame conversions
22:45:32 <nielsm> and the calculation of foundations depends on map data being consistent
22:49:34 <peter1138> Yeah, if it's an object tile it gets called before the object map bits have been moved around.
22:49:51 <peter1138> And st->xy doesn't necessarily point to a station tile.
23:05:07 <peter1138> So I can "fix" the waypoint issue, but I have no idea on the orders.
23:07:26 <peter1138> Travel (for 5 days, not timetabled)
23:07:28 <peter1138> What does that mean?
23:14:05 <nielsm> last time a vehicle completed that leg of the journey it was measured to 5 days
23:17:17 *** andythenorth has left #openttd
23:18:35 <frosch123> peter1138: cargodist needs estimates on how often vehicles complete their route, so it keeps track of how long travel takes even if they times are not timetabled, that is the vehicle does not way when arriving early
23:27:21 <DorpsGek_II> [OpenTTD/OpenTTD] michicc opened pull request #7589: Change: Always apply inflation from 1920 to 2090, no matter the game start year. https://git.io/fjWc7
23:47:29 *** Wormnest has joined #openttd
continue to next day ⏵
