IRC logs for #openttd on OFTC at 2017-03-14
⏴ go to previous day
00:11:31 *** Maarten has joined #openttd
00:24:58 *** HerzogDeXtEr1 has joined #openttd
00:32:02 *** Wormnest_ has joined #openttd
01:01:59 <samu> if (this->state == FLYING) ++this->current_flight_time /= _settings_game.vehicle.plane_speed;
01:03:14 <Wolf01> Use some {} and split it in 3 lines please
01:04:02 <samu> because integer math sucks
01:05:23 <samu> but i need to take plane speed factor into account
01:05:30 <peter1138> I'm not surprised with your made up syntax.
01:07:14 <samu> why is computers doing / so sucky
01:07:47 <samu> i don't get 0.25 + 0.25 + 0.25..., i get 0 + 0 + 0 + 0...
01:08:15 <peter1138> You really have no idea how computers work do you?
01:15:34 <samu> uint64 real_cft = a->current_flight_time / _settings_game.vehicle.plane_speed / _settings_game.vehicle.plane_speed;
01:15:50 <samu> if (this->state == FLYING) ++this->current_flight_time *= _settings_game.vehicle.plane_speed;
01:16:46 <samu> multiplies by 4, only to divide it by 4 twice
01:17:09 <samu> flight time is being done tickly
01:20:31 <samu> ffp dart, with a max speed of 947, goes from right corner to left corner of the map in 3168 *4 / 4 / 4 ticks
01:21:44 <Wolf01> You know that you can even simplify the *4/4 with a *1?
01:22:19 <peter1138> Yes, you should try that.
01:22:37 <samu> gonna try a speed factor 1/3
01:23:55 <samu> i need to count the time with the plane speed factor taken into account
01:24:15 <samu> if i change the speed mid flight, i still want it to count correctly
01:29:31 <Wolf01> What makes you think so?
01:34:01 <samu> if (this->state == FLYING) this->current_flight_time += _settings_game.vehicle.plane_speed;
01:39:05 <Wolf01> But what happen if you change the setting mid flight?
01:46:10 <samu> seems so simple, but i'm stuck
01:50:31 <samu> well, they're all around the same constant value, that's something, but now I'm stuck
01:54:21 <Wolf01> You should also account for diagonal speed
01:56:46 <samu> i changed plane speed mid flight
01:57:10 <samu> current_flight_time 8638 unsigned short
01:57:58 <DorpsGek> Commit by peter1138 :: r27794 trunk/src/network/network_chat_gui.cpp (2017-03-14 01:57:52 +0100 )
01:57:59 <DorpsGek> -Fix [FS#6526]: Chat text background overflowed due to missing padding.
02:00:22 <samu> pff, i'm not sure what I'm doing, guess i better go to bed, get fresh ideas tomorrow
02:01:02 <samu> well, cyas, take care all
02:03:40 <Wolf01> And full moon was 2 days ago
02:05:30 <Wolf01> Btw, I think I'll go to bed too, I want to be here when this madness continues
02:09:50 *** supermop_ has joined #openttd
02:49:41 *** Supercheese has joined #openttd
04:59:00 *** ChanServ sets mode: +v tokai
06:22:33 *** Alkel_U3 has joined #openttd
07:06:16 *** andythenorth has joined #openttd
07:12:27 *** sla_ro|master has joined #openttd
07:48:29 *** andythenorth has left #openttd
07:49:19 *** Hiddenfunstuff has joined #openttd
08:19:49 *** supermop has joined #openttd
08:25:40 *** JacobD88 has joined #openttd
09:08:59 *** FLHerne has joined #openttd
09:21:09 *** Tharbakim has joined #openttd
10:02:25 <Wolf01> I had nightmares after that discussion... XD
10:02:30 <ZirconiumX> (morning everyone)
10:15:08 *** andythenorth has joined #openttd
10:15:33 <Wolf01> Andy, read logs, please
10:15:57 <Wolf01> Such "calc 791 / 1 / 1"
10:16:58 * andythenorth does maths like that too :)
10:17:33 <andythenorth> production / (constant 1 + constant 2)
10:17:57 <Wolf01> But "++x->y /= z" is a shiny diamond
10:20:43 *** maciozo has joined #openttd
10:20:56 <crem> Can you do "int a = 5; ++a = 10; "?
10:21:37 <Wolf01> Even if you can do, it's really difficult to understand at first sight
10:21:59 <ZirconiumX> crem: Not in that context, but x is a pointer in Wolf01's example
10:22:11 <andythenorth> Wolf01: such patches? o_O
10:22:18 * andythenorth is refactoring FIRS
10:22:36 <ZirconiumX> So ++x->y is "add one to the memory address of x and access y"
10:22:46 <crem> I understand the snippet.
10:23:15 <crem> It's indeed lvalue, and "int a = 5; ++a = 10; " compiles.
10:24:35 <ZirconiumX> That sounds like it's borderline undefined due to sequence points
10:24:49 <Wolf01> crem, trust us that's black magic
10:25:10 <crem> In the end? It's unspecified most probably, yes.
10:25:58 <ZirconiumX> According to godbolt.org, GCC without -O emits
10:26:15 <ZirconiumX> Add 1 to a, then a = 10
10:26:23 <ZirconiumX> Resulting in a being 10
10:27:12 <crem> That's how it's logical to work, but I think it's unspecified behaviour because there're two mutations between sequence points.
10:29:26 <ZirconiumX> Even at -O1, GCC converts it to "a = 10"
10:36:59 <andythenorth> if we had deep water, we could let GS control it :P
10:37:22 <andythenorth> I don’t think it adds much except a way for a GS to mess with you :)
10:37:34 <andythenorth> there was a whole edition of Railroad Tycoon with changing sea levels
10:38:09 <andythenorth> peter1138: george has a lot of ideas
10:38:21 <andythenorth> he’s an idea machine
10:38:29 <peter1138> Yeah, but a secret stealthy way to make a vehicle last longer...
10:39:05 <andythenorth> it would be probably on a cargo subtype hack, I assume
10:39:38 <andythenorth> well it’s not the way *I’m* making grfs
10:39:45 <andythenorth> who says I’m right? :P
10:40:01 <Wolf01> BTW, I'm fine with locks for more height levels, but we should have 2 different river slopes, one with rocks and one without, so ships could use the one without
10:41:48 <andythenorth> Wolf01: I am +0.9 to that
10:42:08 <andythenorth> maybe rocks should be a random chance, 1 in 5 or so
10:42:16 <andythenorth> rivers are totally nerfed atm
10:42:20 <Wolf01> Also... falls, different river size based on tile height?
10:43:00 <andythenorth> river generation should be able to make bigger rivers
10:43:16 <Wolf01> Yeah... the current ones are moats
10:43:29 * andythenorth wonders if ‘rapids are 1 in 5 chance’ is an easy patch
10:43:36 <andythenorth> famous last words :P
10:43:47 <Wolf01> Also there is no chance a ship would travel to a mountain peak along a river...
10:43:48 <ZirconiumX> It's easy until you start coding
10:44:11 <andythenorth> probably needs an ‘is rapids’ bit
10:44:18 <andythenorth> and savegame migration and so on
10:44:41 <andythenorth> or a ‘tile is navigable’ bit and no savegame migration :P
10:45:31 <Wolf01> Just make it a setting (only for map generation, like map height)
10:45:55 <Wolf01> Rivers over 16 tile height are rapids
10:46:28 <Wolf01> For old saves it matches the map height, so no river is "converted"
10:47:31 <andythenorth> does ship pathfinder need converted?
10:47:44 <andythenorth> or does concept of navigable / non-navigable slopes already exist?
10:48:09 <andythenorth> peter1138: I’m applying frosch approach: ignore all feature requests ;)
10:48:15 <andythenorth> then they all look valid
10:48:29 <Wolf01> Currently no slope is navigable
10:48:52 <andythenorth> do locks have a bit?
10:48:55 * andythenorth is away from docs
10:49:07 <peter1138> Oh hey, I had a patch for locks didn't I?
10:49:12 <peter1138> But it made ships worse...
10:49:23 <andythenorth> what did it do? o_O
10:49:34 <peter1138> Stops the ships to me them go up and down.
10:50:45 <Wolf01> It just missed some lock animation
10:51:08 <andythenorth> those ships are awful :)
10:51:41 <peter1138> Not really that would require 1) extra graphics 2) making locks accept only 1 ship at a time. And that would make locks far worse.
10:52:24 <ZirconiumX> Is this a case of intentionally being unrealistic to avoid making ships worse?
10:53:21 <peter1138> There was a massive backlash against anything vaguely "realistic" so we don't bother.
10:54:11 <andythenorth> also we’d have to have newgrf graphics
10:54:22 <andythenorth> to allow correct locks for specific locations
10:54:28 <andythenorth> then a choice of locks, like stations
10:54:35 <andythenorth> and a filling-rate property or callback
10:54:40 <andythenorth> and chance of lock breakdown
10:54:49 <andythenorth> and a state machine for correct animation
10:54:54 <andythenorth> and a max ship length
10:55:06 <Wolf01> It doesn't make ships worse, it makes them challenging, also we have all the freedom we want... locks and rivers aren't some original feature and we can change it
10:59:17 <peter1138> I'm pretty sure there was an issue trying to make the water move up & down as well.
10:59:31 <peter1138> And then you might think, oh have multiple locks to speed it up
10:59:39 <peter1138> But then you need to adjust the pathfinder...
11:11:06 <peter1138> Wolf01, that's... awkward.
11:11:25 <Wolf01> Yeah, I'm really impressed
11:11:50 <peter1138> Shame the piano synth is unbearable.
11:12:38 <crem> Undistinguiashable from what a human would generate!
11:13:20 <Wolf01> Computers will replace us someday :D
11:13:40 <peter1138> THE HUMANS ARE DEAD
11:13:57 <peter1138> WE POISONED THEIR ASSES, WITH POISONOUS GASSES
11:14:45 <Wolf01> I would like to know what they will think after some centuries... "why humans died?"
11:15:38 <crem> Registration required to generate, so don't bother.
11:16:57 <Wolf01> This one looks like lego factorio
11:18:06 <samu> how do i get the maximum value that can be set on a setting? i wanna avoid magic number stuff
11:18:18 <samu> _settings_game.vehicle.plane_speed , i know the range is from 1 to 4
11:18:27 <samu> i want to get the max value
11:19:18 <samu> if (this->state == FLYING) this->current_flight_time += 12 / _settings_game.vehicle.plane_speed;
11:20:03 <samu> i wanna change it to 3 * max value of _settings_game.vehicle.plane_speed
11:20:57 <Wolf01> That is a nice way to have a divisible integer.. just make sure it is multiplied by the nearest common multiple
11:21:20 <samu> oh, so it's a bad way to do it
11:21:40 <peter1138> Bah, 140 miles behind target.
11:31:05 <samu> if the cap is changed to 5 in some new version of openttd
11:32:02 <samu> but then 15/5, 15/4, 15/3, 15/2 and 15/1 won't all result as integer
11:33:13 <samu> yeah, 60 is the answer, but how do i calculate that
11:34:20 <samu> how to compute the "divisible integer"
11:40:30 <peter1138> Why does current flight time care about plane speed anyway
11:44:59 <samu> because if the factor is modified, it will travel more tiles in less times
11:45:19 <samu> or less tiles in more time
11:45:52 <samu> perhaps i should rename it
11:47:20 <Wolf01> I would calculate it on plane size
11:48:06 *** minisylf has joined #openttd
12:01:15 <samu> "the least common multiple of consecutive integers"
12:04:10 *** andythenorth has joined #openttd
12:04:11 *** dihedral has joined #openttd
12:36:28 *** chomwitt has joined #openttd
12:43:39 *** supermop_ has joined #openttd
13:38:08 *** sim-al2 has joined #openttd
13:45:34 <Wolf01> Too bad that when they grow up they become brainless idiots
13:55:37 <samu> how do i get the max value of a setting
13:56:18 <samu> something like int x = _settings_game.vehicle.plane_speed
13:56:36 <samu> but for the maximum value that can be set
13:56:46 <samu> not the currently set valuye
13:59:04 <__ln__> oh, microsoft's bing translator translates to/from klingon!
13:59:41 <Wolf01> Samu, do you even try?
13:59:53 <samu> i tried, don't know how to do it
14:00:30 <samu> seems that i need SettingsDesc
14:00:59 <samu> have to access settings.ini
14:01:30 <__ln__> Dun jaj batlhchaj tera' 'ej tlhInganpu'.
14:03:26 <samu> this is what i've came to
14:03:40 <Wolf01> Too bad it doesn't work with visual translation for eastern languages
14:03:46 <samu> for (int i = 1; i <= 4; i++) {
14:03:53 <samu> this line was supposed to be
14:04:19 <samu> for int i = min value of game settings bla bla; i <= max value of game settings bla bla; i++) {
14:04:36 <samu> how do i retrieve those values :(
14:04:52 <peter1138> Do you think it will ever change?
14:04:53 <Wolf01> You said that, SettingDesc
14:04:59 <LordAro> samu: generally, by searching through the code for existing examples
14:05:19 <LordAro> evidently no one else knows
14:06:14 <samu> i know it's going from 1 to 4, but it can change some day, i dunno
14:06:21 <peter1138> More like nobody else wants to do something so hideous.
14:09:37 <LordAro> samu: however, it looks like you want GetSettingFromName
14:24:30 <andythenorth> that’s standard UI programming :P
14:24:34 <andythenorth> it’s always a hack
14:40:36 *** supermop has joined #openttd
14:57:27 <supermop> man work vpn is so slow
15:03:49 <supermop> andythenorth: is 2cc trams can't be persuaded, and zeph is missing
15:04:25 <supermop> I guess next step is i try to add more nrt stuff in a ogfx+ add on?
15:04:48 <supermop> there are a couple abandoned trolleybus sets out there
15:05:23 <planetmaker> supermop, I guess, with proper version query, this can be simply added to stock ogfx+
15:05:58 <planetmaker> though not entirely sure... would need to check how to find out the openttd version in a non-linear way, I guess
15:14:57 <supermop> sounds like the solution is to just go ahead and squeeze nrt in before the 1.7 release in two weeks
15:15:38 <samu> uint plane_speed_max = GetSettingFromName("vehicle.plane_speed", &plane_speed_max)->desc.max;
15:16:14 <andythenorth> planetmaker: for Road Hog I just branched
15:16:17 <samu> if (this->state == FLYING) this->current_flight_time += CommonWhatever(plane_speed_max) / _settings_game.vehicle.plane_speed;
15:16:26 <andythenorth> jenkins hates me, because I had to specify an nml path on my local OS
15:26:04 <ZirconiumX> How would you even go about that?
15:26:31 <ZirconiumX> Unreleasing something
15:27:12 <Wolf01> <supermop> sounds like the solution is to just go ahead and squeeze nrt in before the 1.7 release in two weeks <- cool but no... to many missing features, I don't want a bunch of awesome grfs which need to be reworked when road sidings and other flags will be added
15:29:04 <planetmaker> andythenorth, branched what? OpenTTD?
15:29:20 <planetmaker> or that newgrf to get the nrt stuff in? I guess that :)
15:29:28 <planetmaker> I guess branching is the easiest, yes
15:29:29 <andythenorth> planetmaker: I branched Road Hog yes ;)
15:29:49 <andythenorth> jenkins has no way to use the forked nml compiler though
15:30:06 <planetmaker> hm... forked nml... that's more difficult yes
15:30:38 <planetmaker> it has ways to use it, but it needs probably ssh to the server
15:30:43 *** frosch123 has joined #openttd
15:30:47 <planetmaker> in order to setup that stuff
15:31:50 <planetmaker> andythenorth, where's the nml fork?
15:32:57 *** Ethereal_Whisper has joined #openttd
15:34:21 <planetmaker> that definitely excludes it from adding it quickly, sorry
15:34:50 <andythenorth> it could be pushed to devzone
15:35:05 <andythenorth> it’s fine for me, I can compile locally :)
15:35:10 <andythenorth> but jenkins has angry red
15:35:16 <andythenorth> maybe I should disable builds on the branch
15:36:12 <andythenorth> oh actually maybe it’s python failing there
15:36:35 <andythenorth> when it starts going continuosly red, errors get ignored :)
15:37:31 <planetmaker> you always break stuff. That's why I'm not concerned about red on your projects :P
15:38:05 <andythenorth> I try to keep them green, but eh
15:38:25 <andythenorth> I don’t read email, and I’m not much in coop channel, so I miss the warnings
15:39:11 *** Ethereal_Shiver has quit IRC
15:39:15 <planetmaker> it anyway dwindled down to a build log
15:39:27 <planetmaker> and thus a status log of your builds :P
15:54:07 *** Ethereal_Shiver has joined #openttd
15:58:55 *** supermop_home has joined #openttd
15:59:36 *** Ethereal_Whisper has quit IRC
15:59:47 <andythenorth> so reproing OS X bugs like this one….
16:02:11 <andythenorth> same hardware? o_O
16:02:14 <frosch123> you need also an ottd server to connect to which reproduced the same event sequence
16:02:22 <andythenorth> so it’s never happening eh? o_O
16:02:32 <frosch123> on the plus side, i don't think it would be osx related
16:02:33 <andythenorth> unless we somehow virtualise the whole world, and do it in a simulation
16:02:53 <andythenorth> actually simulated reality would kick ass for bug fixing
16:03:02 <andythenorth> we should work on that
16:03:10 *** supermop_home_ has joined #openttd
16:03:22 <frosch123> step 1: get rid of all human players?
16:04:23 <frosch123> step 2: clarify whether that includes V
16:05:24 *** Alberth has joined #openttd
16:05:24 *** ChanServ sets mode: +o Alberth
16:05:49 <andythenorth> step 2 takes a while
16:06:15 <frosch123> step 2 is always the difficult part
16:06:18 <frosch123> but step 3 is profit
16:07:06 <andythenorth> is it a tube race __ln__ ?
16:07:12 * andythenorth didn’t watch to end
16:07:31 <andythenorth> think my brother tried that
16:12:40 <andythenorth> if FIRS wasn’t 483279 lines of nml, would it compile faster? o_O
16:15:26 <Alberth> delete the empty lines :p
16:15:36 <ZirconiumX> andythenorth: computers boot faster when they have less to load, so if you wipe your hard drive they boot immediately
16:15:36 <Alberth> and all double white-space :p
16:16:11 <Alberth> my first computer booted from ROM, now that was fast :p
16:18:21 *** supermop has joined #openttd
16:18:59 <frosch123> old computers were also really fast to turn off :p
16:19:21 <Alberth> my fedora gnome is quite fast with that, tbh
16:20:07 <Alberth> now it would be useful if it could also open a gui window :p
16:20:22 *** supermop_ has joined #openttd
16:23:11 *** supermop_home_ has quit IRC
16:32:34 *** sla_ro|master has joined #openttd
16:34:16 <peter1138> There was never a use for fast ships, was there?
16:34:33 *** supermop_home has joined #openttd
16:34:50 <crem> Before 20th century there were.
16:34:50 <Alberth> well, on occasion, people want race ships :p
16:35:16 <crem> Actually yes, for entertaiment for sure.
16:35:17 <Alberth> speedy delivery of them mailbags
16:35:20 <peter1138> Yeah but... 80km/h+
16:35:42 <peter1138> (Convert to knots as pedantic)
16:36:04 <Alberth> don't you know realism limits apply selectively?
16:36:35 <peter1138> More like adding complexity for no reason
16:37:03 <Alberth> that too, but not a player consideration :)
16:46:17 *** Wormnest has joined #openttd
16:47:39 <frosch123> peter1138: i believe people tried to make ecranoplanes as ships
16:48:13 *** andythenorth has joined #openttd
16:54:55 <samu> i'm undecided if I store this variable as uint16 or uint32
16:57:09 <crem> That's not extremely realistic.
16:57:25 <crem> Also, probably passengers are a bit bored after 221 days.
16:57:39 <crem> But they surely know all each other well.
16:59:54 <samu> @calc 4294967295 / 74 / 4
17:00:02 <crem> @google Calories in 400 tons of butter.
17:01:36 <samu> is there a middle container, something between uint16 and uint32?
17:02:06 <Alberth> you can of course only use 20 bits of the 32 :p
17:03:42 <Alberth> don't worry about that single byte, it's not worth the trouble
17:04:03 <samu> isn't uint32 something quite too large to be computed daily for all aircraft?
17:04:47 <crem> 32-bit arithmetics are the same, if not faster, than 16-bit one.
17:05:26 <crem> For x64 the same is true for 64-bit numbers.
17:05:26 <Alberth> the real saver in performance would be not to increment ticks each tick
17:05:50 <crem> How many ticks are there per second?
17:06:21 <Alberth> crem: there is the L1/L2 cache issue :p
17:06:47 <Alberth> longer numbers use more cache
17:08:01 <andythenorth> oh that was said already
17:08:04 <andythenorth> trolling fail :)
17:08:55 <crem> uint24 would be slower than either uint16 and uint32 on many architectures due to alignment.
17:09:10 <crem> (need to read 2 machine words in some cases rather than 1)
17:09:20 <samu> i think 221 days in flight can occur on very large maps
17:09:23 <frosch123> 20 years ago, graphics card had uint24 :)
17:09:38 <frosch123> and floppys had fat12
17:09:51 <Alberth> if you have a few million aircraft, 24bit makes sense
17:10:05 <crem> It's memory usage vs of performance tradeoff.
17:10:19 <crem> And samu was worried about performance.
17:11:00 <Alberth> save the start time, and performance ceases to be a problem
17:11:57 <Alberth> samu: what is supposed to happen?
17:12:14 <samu> there's plane speed factor which can be changed at any time
17:12:27 <Alberth> is $random long flight time to know exactly to the tick?
17:13:03 <Alberth> ie does it make a difference beteen 221 and 220 days, eg
17:13:16 <Alberth> or is "long" enough to store?
17:13:46 <samu> probably doesn't make a difference, i could try counting it every day, just wanted to be as thorough as possible
17:14:24 <samu> but then changing the plane speed factor might affect the result slightly
17:14:44 <Alberth> it's less flying or something?
17:15:37 <samu> sec, let me copy paste patch
17:15:38 *** supermop has joined #openttd
17:17:13 <samu> this is what i have so far... just noticed i mixed lifetime profit patch with it, lol
17:17:30 <samu> ignore the lifetime profit stuff, it's not mine
17:19:41 <samu> line 145 is to be edited yet, i have not finished it
17:19:55 <peter1138> yeah well what are you even doing with this flight time thing?
17:20:27 <samu> if it goes past a certain value
17:20:35 <samu> which i have yet to figure out
17:22:06 <peter1138> and you want it to be affected by plane speed?
17:22:29 <Wolf01> I tried to tell him to calculate it based on plane size :(
17:22:34 <samu> only the plane speed factor
17:23:11 <peter1138> so when plane speed makes planes slow, they can last longer? or less time?
17:23:16 <Alberth> start with the simplest possible form of counting, then figure what you actually want
17:23:53 <samu> basically they will travel the same distance before the plane enters breakdown, no matter the plane speed factor that is set
17:24:10 <samu> at least that's what i was trying to achieve
17:24:46 <peter1138> why not just use distance?
17:25:06 <samu> which distance? between stations?
17:25:37 <samu> because aircraft can be ordered to go somewhere else mid flight
17:28:20 *** supermop_home has joined #openttd
17:28:59 <samu> also, aircraft travel faster on axis than diagonal, wouldn't be accurate
17:29:17 <LordAro> accuracy is not something ottd's known for
17:29:21 <andythenorth> so I’ve got #includes in my generated nml
17:29:26 <andythenorth> but they don’t exist in my repo :P
17:29:29 <Wolf01> Livery also changes air drag, red is fast
17:33:04 <andythenorth> file was open in text editor, unsaved to disk
17:33:45 <andythenorth> editor has a multiple-files search, but that looks at editor state of any open files, not filesystem state
17:34:03 <andythenorth> I thought the editor was 100% using the disk
17:34:14 <andythenorth> as it has to select a path :P
17:35:35 <samu> trying to figure out a "max_distance_before_forced_breakdown" value now
17:36:11 <samu> not really distance distance
17:36:23 <samu> distance with plane factor accounted
17:36:47 <samu> i dunno what to call the variable, but it's of that nature
17:38:13 <peter1138> Hmm, I guess that affected everyone's task list on flyspray...
17:38:41 <peter1138> Added the last edited column, but it seems to be project wide not per user.
17:39:14 <Wolf01> Yes, it's project wide
17:39:35 <Wolf01> But IIRC there's something for users too
17:40:35 <samu> currently sending an helicopter accross a 4096x4096 map in debug mode... zzzzz
17:40:59 * LordAro hears TB screaming in the background
17:42:11 <samu> wanna see its final flight_counter value
17:42:52 <peter1138> Hmm, better not commit all these Console.WriteLines...
17:43:21 *** TheMask96 has joined #openttd
17:43:26 <samu> why is debug mode so slow :(
17:43:31 <andythenorth> debugging innit :)
17:45:50 <samu> gonna try release mode, i can't wait this long, hope i can still read value
17:47:02 *** supermop has joined #openttd
17:52:21 <andythenorth> just 3 #defines left in FIRS
17:52:27 <andythenorth> what is done, is not always easily undone :P
17:53:05 <andythenorth> and this is how we learn about separating data from presentation
17:53:17 <andythenorth> although /me already knew about that, and is removing other people’s work :)
18:00:35 <DorpsGek> samu: Error: Something in there wasn't a valid number.
18:01:19 <samu> @calc 2*2*2*2*2*2*2*2*2*2*2*2*2*2*2*2*2*2*2
18:02:46 <LordAro> why does that matter?
18:02:54 <LordAro> that would be ridiculously inefficent to implement
18:03:26 <andythenorth> isn’t there something about allocating memory in 8s? :P
18:03:30 <Alberth> make a 1 bit computer, and hook 19 of them after each other
18:03:32 * andythenorth is not a programmer
18:04:36 <Alberth> transputer did something in that direction iirc
18:05:18 * LordAro needs to write some stuff about that at some point
18:05:49 <samu> well, i'm probably changing it to uint16, limit would be around ~221 days in flight
18:06:12 <samu> or maybe i should count it daily instead of tickly
18:07:30 <LordAro> you're allowed to use an integer bigger than you need
18:09:56 <andythenorth> eventually became XMOS, near my office :)
18:10:17 <andythenorth> I know David May, the co-inventor of the transputer, although only approximately :)
18:10:47 <Alberth> doesn't seem to talk about 1 bit computers that you connect
18:11:10 <Alberth> although it had very advanced multi-tasking capabilities
18:12:12 <Alberth> synchronous communication between processes, in hardware
18:15:51 <andythenorth> yeah can’t find any 1 bit things there :)
18:18:49 <andythenorth> Alberth: the CPP is nearly gone from FIRS…that clears the way for a....
18:18:55 <andythenorth> …new makefile :D
18:20:10 <Alberth> not sure how makefile and not running cpp is connected, but \o/ :)
18:21:35 <andythenorth> therefore easier replacement
18:22:50 <Alberth> computer with lots of blinking lights always wins :)
18:23:40 <andythenorth> says they had to blink the lights deliberately, for effect :)
18:26:19 *** HerzogDeXtEr has joined #openttd
18:42:52 <samu> flight_counter 419853 unsigned __int64
18:43:16 <samu> crossed a map from right to left, 4096x4096 tiles
18:46:11 <samu> much higher than the cap of ~221
18:47:06 <samu> yeah, think it's time to store this in days insted
18:51:26 <andythenorth> threw my backup drive across the room :P
18:52:29 <ZirconiumX> Not much of a backup now
18:53:57 <LordAro> do you have a backup backup?
18:58:17 <andythenorth> I should be on telly
18:58:21 <andythenorth> that’s how good I am
18:59:15 * peter1138 ponders trying to fix this helicopter issue
18:59:47 <andythenorth> SSD backup was kind of an accident, not because I bleed gold when I cut myself :)
19:00:17 <peter1138> That's a shame, I was going to tap an artery.
19:00:28 <andythenorth> bought a new laptop 3 years ago, pulled the SSD from the old laptop to transfer files over…
19:00:38 <andythenorth> put it in a case, voila, SSD backup :P
19:02:27 * andythenorth wonders how much SSD is these days
19:02:31 <andythenorth> long time since I bought one
19:02:48 <peter1138> Not a lot for the smaller ones.
19:03:28 <andythenorth> a shitload for the bigger ones
19:03:31 * andythenorth looking on crucial
19:04:53 <andythenorth> yeah and no, I’ll keep my 1TB spinning disk other backup
19:33:46 <frosch123> i wondered, should we rename opntitle.dat to opntitle.sav? just so people do not have to ask how to load the titlegame?
19:37:37 <Eddi|zuHause> no, that would make too much sense
19:39:45 *** chomwitt1 has joined #openttd
19:45:45 <DorpsGek> Commit by translators :: r27795 trunk/src/lang/luxembourgish.txt (2017-03-14 19:45:36 +0100 )
19:45:47 <DorpsGek> luxembourgish: 3 changes by Phreeze
19:50:45 <andythenorth> frosch123: and why not? :)
19:51:00 <andythenorth> renaming things is a feature
19:51:46 <frosch123> i can't tell whether that is a yay or nah
19:52:22 <andythenorth> .dat means nothing to me
19:52:22 *** Stimrol has joined #openttd
19:55:18 *** Progman has joined #openttd
20:00:23 <ZirconiumX> For a dedicated server, does libLZMA matter?
20:00:29 <Rubidium> frosch123: just imagine what the "reverse" effect would be... questions whether OpenTTD crashes upon starting due to NewGRFs in the intro game
20:00:49 <frosch123> does no longer happen
20:01:03 <frosch123> if there are newgrf in the intro game, it goes to water-only
20:01:26 <frosch123> it's a hidden but intentional side-effect of the town-newgrf fix :)
20:01:56 *** gelignite has joined #openttd
20:02:02 <Rubidium> well, then: why it doesn't load the savegame I named to opntitle.sav?
20:02:06 <Eddi|zuHause> ZirconiumX: no, only for loading specific savegames
20:02:47 <frosch123> possible, we can easier blame people who rename their files into .dat :)
20:02:53 <Eddi|zuHause> ZirconiumX: however, lzma2 (or xz) provides the best (tm) compression
20:03:11 <ZirconiumX> Which is why new savegames use lzo2 :p
20:03:16 <Rubidium> ZirconiumX: only caveat is that lzma is the best compromise between time and size for the compression of savegame in case of network transfer
20:03:42 <samu> make it multi-threaded ploz
20:03:50 <ZirconiumX> I ask because I'm cross-compiling for my raspberry pi (don't ask)
20:04:10 <ZirconiumX> So building liblzma and liblzo2 purely for this would be awkward
20:06:31 <Rubidium> samu: that's not really helping. The most "wasted" time is in cloning the game state and network transfer; after the cloning compression starts in a separate thread and whenever enough bits are compressed to send to the client, they are sent to the client. From that moment on it's mostly network IO
20:06:40 <Eddi|zuHause> ZirconiumX: you might have trouble getting things from bananas if you skip essential compression algorithms
20:06:41 <Rubidium> that's causing the delays
20:07:13 <ZirconiumX> Eddi|zuHause: curl <blah> | tar
20:07:40 <Eddi|zuHause> not what i meant
20:08:46 <Eddi|zuHause> ZirconiumX: with curl/wget you can't ask bananas for specific not-newest versions of files
20:09:24 <Eddi|zuHause> whereas the ingame download can handle that
20:09:38 <ZirconiumX> This is for a dedicated server
20:10:06 <ZirconiumX> So bananas is irrelevant for testing
20:10:12 <Eddi|zuHause> you make a savegame on your client, upload it to the server, and tell it to fetch the dependencies
20:14:41 <ZirconiumX> Wasn't there a flag to set the version of a compile?
20:14:51 <ZirconiumX> --with-version=BLAH or something?
20:15:08 <Eddi|zuHause> even if there was, you should not use it
20:15:21 <frosch123> it was burned at high temperature
20:15:39 <ZirconiumX> Okay, so I just have to live with OpenTTD being very well designed.
20:15:46 <frosch123> after coop used to configure completely random revisions and then complained about desyncs
20:17:07 <ZirconiumX> I have a client compile versioned "jgrpp-0.17.2", and a server compile versioned "jgrpp-0.17.2-14" and the two won't talk to each other.
20:17:16 <ZirconiumX> Despite being built from the same source code.
20:18:04 <ZirconiumX> server is "jgrpp-0.17.2-44-gd14d573"
20:18:48 <Eddi|zuHause> everything before the g<xxx> is the branch/tag name
20:20:37 *** ChanServ sets mode: +v tokai
20:49:11 *** supermop__ has joined #openttd
20:55:10 <andythenorth> gah backups are boring
20:55:28 <andythenorth> 200GB takes 2 hours :(
20:55:40 *** taco is now known as Guest320
21:03:32 * andythenorth teddybear moaning
21:04:53 *** Supercheese has joined #openttd
21:11:25 <Eddi|zuHause> generally, if something takes 2 hours, maybe you should be looking for something else to do... watching it will make it take longer
21:12:10 <frosch123> he was looking into moaning about it
21:16:20 <Eddi|zuHause> you know what i hate about python? that you quit the interactive shell differently on windows and linux
21:16:46 <frosch123> never used it on windows
21:17:29 <Eddi|zuHause> on linux you type ^D and on windows you type ^Z[enter]
21:18:31 <frosch123> well, just use ^Z on linux as well
21:18:39 <frosch123> and get used to type "fg" to restart it
21:18:50 <Eddi|zuHause> well, that doesn't really "quit" :p
21:18:53 <andythenorth> Eddi|zuHause: I have many other things to do :)
21:19:15 <frosch123> "have to do" or "doing"?
21:19:32 <andythenorth> if you don’t have kids, bed time is hard to explain :P
21:19:36 <andythenorth> much tears today
21:20:07 <andythenorth> child #2 wanted me to continue reading the snake book, and is in tears because I stopped before the boa constrictor page
21:20:20 <andythenorth> also he trapped his finger in a door and has a big piece of skin hanging off
21:20:37 <andythenorth> child #1 is learning swift (Apple’s fake javascript or whatever it is)
21:21:10 <andythenorth> and is in tears because the tutorial is teaching him how to nest common code into repeatable functions, and he just wants to repeat his commands long-hand
21:21:16 <frosch123> when i am reading the favorite books of my nice and nephew, i usually alter the story incrementally, until they notice :)
21:22:20 <andythenorth> it’s a good tactic
21:22:28 <andythenorth> gets interesting when they can actually read
21:22:32 <frosch123> usually they are upset :)
21:23:03 <Eddi|zuHause> reminds me of my (younger) sister, she used to be bad at reading, so when reading comics she just looked at the pictures and made up a story for them, instead of reading the speech bubbles
21:23:18 <frosch123> nah, i mean that kind of books which they know every word of, and could "read" without being able to "read"
21:23:57 <Eddi|zuHause> my (even younger) brother always corrected her
21:24:13 <frosch123> at some point they want people to read the books to them, but are actually too bored to listen since they know it already
21:24:31 <frosch123> then i troll them by altering the story
21:26:04 <Eddi|zuHause> my nephew has a CD player, and when he wants to restart the disk, he doesn't use the stop/play button, but he opens it up and closes it again (it will then auto-play)
21:26:06 <andythenorth> just pushed the last removal of CPP from FIRS
21:26:48 <frosch123> Eddi|zuHause: he, i was once looking at old photos with my nice, and pointed out my sister on them by saying "see, that is <name of sister>". but then my nice insistet on the person being her mother :p
21:28:11 <Eddi|zuHause> frosch123: there exists a picture of my father as a kid, and when i was that age, my parents showed that picture to me and my siblings, and everybody agreed that it was me on that picture
21:28:51 <frosch123> yep, i also know some son/father pairs like that. esp. funny if your put photos of them next to each other
21:31:15 <andythenorth> child #2 looked exactly like me when he was a baby
21:31:24 <andythenorth> but…not like I did when I was a baby :P
21:31:45 <andythenorth> have we closed any more bugs? o_O
21:31:52 <andythenorth> I tried to make sense of the OS X bugs, but eh
21:32:14 <frosch123> shall i open more for you to fix?
21:33:52 <andythenorth> I think you are playing a different game :P
21:34:05 <andythenorth> my game is always to be left with a handful of issues / tickets :P
21:36:31 <frosch123> well, i create tickets for bugs which i hope other would fix
21:36:50 <frosch123> i don't care about creating tickets for myself
21:37:17 <andythenorth> those chips tickets, I am hoping yexo will come back and fix :P
21:37:22 <andythenorth> it’s been a while :)
21:38:31 <frosch123> hmm, maybe i could use chips as testcase
21:39:15 <frosch123> i was considering to start nml from the other end
21:39:28 <frosch123> and write a grf linker
21:39:57 <frosch123> which takes multiple intermediately compiled input files, and links them into a grf
21:40:42 <andythenorth> for faster, or better in other ways?
21:41:22 <frosch123> separation of abstraction layers and stuff
21:42:16 <andythenorth> Alberth helped, but don’t blame him for my ideas, it wasn’t his fault
21:42:33 <andythenorth> I was looking at the declaration end
21:43:23 <frosch123> i know, but i wanted to start at the other end
21:43:33 <frosch123> since that appeared to be least explored
21:43:33 <andythenorth> I can see why :)
21:43:55 <andythenorth> I could pretty much test my ideas with some horrible python, it’s a low skill requirement
21:54:21 *** Hiddenfunstuff has quit IRC
21:54:35 <Alberth> :o more nml movements :)
22:09:05 *** sim-al2 has joined #openttd
22:11:46 <samu> v->flight_counter = ClampU(v->flight_counter + m / _settings_game.vehicle.plane_speed, 0, UINT16_MAX / speed_factor_max);
22:12:05 <samu> is ClampU going to do what I'm thinking it's gonna do?
22:13:14 <samu> if the value exceeds 16383 and flight counter is above that value, set it to always be 16383
22:15:29 <samu> flight_counter is being checked daily
22:15:58 <samu> it can check for a max of 16383 days then
22:16:12 <samu> because flight_counter is a uint16
22:17:51 <Alberth> oh, just 16K-ish, not even near any limit
22:20:48 <samu> flight_counter is 4 times that value
22:23:17 <samu> m / _settings_game.vehicle.plane_speed
22:24:18 <samu> i must be doing something wrong, grrr... brb
22:33:36 <samu> 1 day at speed factor 1/4 counts as 3 per day
22:34:02 <samu> 1 day at speed factor 1/3 counts as 4 per day
22:34:12 <samu> 1 day at speed factor 1/2 counts as 6 per day
22:34:20 <samu> 1 day at speed factor 1/1 counts as 12 per day
22:35:50 <samu> 5461 days, that's a bit less than i initially thought
22:38:41 *** Progman_ has joined #openttd
22:39:36 <supermop> tired from trudging in snow
22:40:33 *** Stimrol has joined #openttd
22:41:59 *** Progman_ is now known as Progman
22:46:20 <supermop> don't have to energy to comment in DOO thread
22:49:34 *** FLHerne has joined #openttd
22:59:14 <samu> it's correct after all, i was just being dumb, it's sad how I forget why I did the things the way I did
22:59:48 <samu> i will have a sad life when i get older
23:00:30 <Eddi|zuHause> what's a DOO thread?
23:20:12 <supermop> driver only operation
23:24:25 *** FLHerne has joined #openttd
23:59:06 *** Arveen2 has joined #openttd
continue to next day ⏵
