IRC logs for #openttd on OFTC at 2019-01-23
⏴ go to previous day
00:00:07 <Samu> find options, use regular expressions, build expression, hmm and now what to pick?
00:11:37 *** Thedarkb1-T60 has joined #openttd
00:29:25 <Samu> that was easier than expected
00:29:40 <Samu> STR_AI_CONFIG_HUMAN_PLAYER.*
00:31:19 <Samu> it left an empty line... i fail
00:43:41 <Samu> STR_AI_CONFIG_HUMAN_PLAYER.*\w\r?\n
00:46:26 <DorpsGek> planetmaker: I have not seen adf.
01:00:53 <glx> Samu: maybe too late but ^.*STR_AI_CONFIG_HUMAN_PLAYER.*$ should get the whole line
01:07:19 <glx> you used the export scripts ?
01:16:20 <Samu> export scripts? no, what is that?
01:17:13 <Samu> Edit > Find and Replace > Replace in Files
01:17:20 <Samu> went here on visual studio
01:18:18 <glx> I mean squirrel_export and generate_widget
01:30:43 <glx> using the scripts I get the same src/script/api changes than your manual edits
01:31:11 <glx> (generate_widget is not runnable in current master)
01:32:53 <glx> but you can run squirrel_export.vbs yourself, it's better to use "cscript squirrel_export.vbs" if you don't want to be assaulted by message boxes in case many changes happen
01:48:47 *** raphikov has joined #openttd
02:01:50 *** Thedarkb-X40 has joined #openttd
02:40:29 *** snail_UES_ has joined #openttd
03:03:52 *** Wormnest has joined #openttd
03:17:48 <DorpsGek_II> [OpenTTD/OpenTTD] nikolas opened pull request #7086: Change #6173: Update SDL driver to use SDL 2.0 https://git.io/fhamZ
07:41:06 <peter1138> But "^.*STR_AI_CONFIG_HUMAN_PLAYER.*$" is the same as... "STR_AI_CONFIG_HUMAN_PLAYER" ...
08:02:44 <peter1138> TrueBrain, will there be Linux binaries for nightlies?
08:27:41 *** andythenorth has joined #openttd
08:41:09 <andythenorth> they allowed Pikka back
08:41:23 <Pikka> time off for good behaviour, or something
08:41:27 <andythenorth> how's pineappple land? 40 degreees?
08:41:31 *** HerzogDeXtEr has joined #openttd
08:41:55 <andythenorth> dunno if I'm jealous or not
08:42:10 <andythenorth> it's 3 here and dark
08:42:37 <peter1138> -0.3 at the moment.
08:43:27 <andythenorth> Pikka: so what's the go? :)
08:44:31 <Pikka> some of each, I think. Internships for uni this year, which could be fun.
08:45:10 <andythenorth> column A, column B
08:45:18 <andythenorth> hmm why did I delete HSTs?
08:45:32 <Pikka> something about mail capacity
08:46:17 <Pikka> I didn't put mail capacity in mine, let people use them on coal trains if they want to ;)
08:46:30 <andythenorth> nah, it's because I have an 8/8 pax engine already with ~3000hp
08:46:42 <andythenorth> so a 16/8 HST engine at 4500hp wins nothing
08:47:11 <andythenorth> that could change, but pulling that thread....changes everything else :P
08:47:26 <andythenorth> anyone know if Horse is done yet?
08:47:58 <Pikka> yes, adding a delticoid will blunt the HST somewhat
08:48:30 <andythenorth> I could get over my obsession with integer lengths
08:48:38 <andythenorth> and do a 12/8 HST
08:51:15 <Pikka> at least with real-world vehicles, I can blame redundancy on "realism" :)
08:52:44 <Pikka> also I decided to be stricter about 20th century, so the Voyager is out and the HST becomes an end-game vehicle :P
08:53:32 <andythenorth> you need more bi-mode :P
09:56:53 *** andythenorth has joined #openttd
10:08:15 <peter1138> Hey Pikka, have you considered not deleting all your website content every couple of months? :/
10:08:47 <andythenorth> serves faster though
10:11:15 <andythenorth> imagine the speed of it
10:23:07 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN updated pull request #7077: Change: Always build baseset metadata, instead of including it in the repo. https://git.io/fhzn8
10:24:16 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN commented on pull request #7077: Change: Always build baseset metadata, instead of including it in the repo. https://git.io/fhaEd
10:24:45 <peter1138> Azure being shit again :p
10:25:30 <peter1138> anybody knows why #7077 - my stuff that wasn't build?
10:28:22 *** tokai|noir has joined #openttd
10:28:22 *** ChanServ sets mode: +v tokai|noir
10:33:52 <peter1138> Hmm, should I crack on with work, or look at #6811? :p
10:36:41 <Pikka> I thought I was going to do something else with the website
10:37:15 <andythenorth> peter1138: bit of column A, bit of column B?
10:37:16 <Pikka> so I put the previous last post back :)
10:37:22 <andythenorth> how urgent / fun is the work? o_O
10:40:13 <andythenorth> Pikka: signs Horse might be done...
10:40:24 <andythenorth> I spent a day making EMUs go faster if you attach the right engine
10:42:04 <Pikka> wasn't there something about redrawing all the steam locomotives? ;)
10:42:12 <andythenorth> oh yes, they're not done
10:42:33 <andythenorth> also I considered 'brake force', but it turns out Snail already did it
10:42:54 <andythenorth> I even considered....variable running costs :o
11:11:58 <peter1138> We need brake force in game :/
11:12:20 <peter1138> I just have no idea how to properly implement it :p
11:13:06 <peter1138> Hmm, work out how many tiles it'll take to stop at the current speed, and then reserve those extra tiles. Somehow.
11:13:50 <peter1138> If (reservation length < braking length) try reserving further, else start slowing down
11:13:55 <peter1138> Hmm, that may actually work.
11:14:16 <peter1138> It would also make path signals stay green longer.
11:15:06 <peter1138> Not sure about slopes.
11:16:47 <andythenorth> can we have acceleration (torque) *and* deceleration (brake force)? :P
11:16:56 <peter1138> I'd approve #7085 but Azure hasn't run that either.
11:17:07 <peter1138> andythenorth, please!
11:17:15 <peter1138> andythenorth, but is there a sensible default for torque?
11:17:27 <peter1138> Is it currently "infinite"?
11:17:44 <andythenorth> can't remember, let me look in ground_vehicle :P
11:18:00 <peter1138> I think it would be, as it's just mass and max_te.
11:19:02 <peter1138> Would it be an absolute value or a coefficient?
11:19:07 <andythenorth> coefficient I think
11:19:12 <andythenorth> I looked into real physics of it, but eh
11:19:22 <andythenorth> it just needs to be a bonus / malus for vehicles
11:19:25 <andythenorth> I've tested in-game, and TE factor is pretty pointless
11:19:46 <andythenorth> TE only makes a gameplay difference on slopes where the vehicle is already down near lowest speed
11:19:53 <andythenorth> this is correct physics afaict also
11:22:13 <andythenorth> gameplay result is that given a choice of similar engines, you should always pick the engine with most HP
11:45:23 <peter1138> Our slopes aren't long enough
11:45:37 <peter1138> It makes a difference when a heavy train is on a slope
11:45:59 <peter1138> This is why the freight trains factor exists, to artifically inflate the gradient.
11:46:25 <andythenorth> I tested it quite a bit, it's only a factor at lower speeds
11:46:32 <peter1138> Maybe it would be better to apply to the whole train, not just the weight of freight
11:46:35 <andythenorth> matches the real physics also
11:46:45 <peter1138> Torque is also such a factor
11:46:49 <andythenorth> once TE is 'enough', HP is the major factor IRL
11:47:00 <andythenorth> real vehicles have limitations like rotation speed of components
11:47:12 <andythenorth> I don't want to model that
11:47:15 <peter1138> That's why we have a max-speed property
11:47:24 <andythenorth> I just want the option to nerf the acceleration of fast trains
11:47:40 <peter1138> So torque to limit acceleration.
11:47:48 <andythenorth> IC125 accelerates slowly :P
11:47:51 <andythenorth> due to high gearing
11:47:56 <peter1138> If it's a coefficient, what is it a coefficient of?
11:48:19 <andythenorth> dunno, I tried to figure it out, irl the physics equations don't account for it
11:48:24 <peter1138> If it's an absolute value, that kinda makes more sense and easier to fit in.
11:48:39 <andythenorth> because HP is just derived from torque * speed / 5252 or something
11:50:07 <andythenorth> Horsepower = (Torque x RPMs) / 5252
11:51:04 <peter1138> We don't simulate RPMs of course.
11:51:09 <andythenorth> well we have speed
11:51:28 <andythenorth> which is related to RPMS by pi* r^2 I guess
11:51:29 <peter1138> Should we simulate gearing?
11:51:55 <andythenorth> does your 29-er get off the line faster than your BMX?
11:52:02 <andythenorth> given the same peter1138 on the pedals?
11:52:19 <peter1138> I don't have a BMX.
11:52:29 <peter1138> So I can't tell you.
11:52:55 <andythenorth> I never understood why
11:53:00 <peter1138> Could you simulate gearing by using different max torques at different speeds?
11:53:08 <andythenorth> I don't understand the IRL maths
11:53:17 <andythenorth> the smaller wheel is easier to spin, but travels less far
11:53:18 <peter1138> I'm not sure that'd work.,
11:53:28 <andythenorth> so why doesn't the maths net out to same result?
11:53:35 <andythenorth> is it because r^2 ?
11:53:36 <peter1138> Because it's not just the wheel.
11:53:39 <andythenorth> where is Eddi|zuHause :P
11:54:16 <andythenorth> isn't work done / time going to come out the same at the end?
12:33:20 <Eddi|zuHause> i'm not sure what you're trying to get here
12:34:29 <andythenorth> 1. understanding
12:34:45 <andythenorth> 2. more ability to differentiate low speed / high speed engines
12:35:53 <Eddi|zuHause> the max rotation speed of the motor comes usually from other factors, so the gear ratio directly translates that into a max vehicle speed
12:36:42 <andythenorth> those factors aren't captured by f = ma or HP = (torque * rpm) / 5252
12:37:54 <Eddi|zuHause> at low speeds speed is limited by max TE (wheels slipping on the rails), and at higher speeds it is limited by power/speed - friction > 0
12:38:35 <andythenorth> - electric motors produce max torque at stall, and are limited on rotation speed, primarily by back EMF
12:38:49 <andythenorth> - diesel engines produce more torque at higher speeds, but have a fixed RPM range
12:39:06 <andythenorth> - steam engines with cylinders are limited by the rotational speed of the wheel bearing and other motion parts
12:40:12 <Eddi|zuHause> that needs a more detailed physics emulation
12:40:36 <andythenorth> well, sticking to 'how does RL work'...
12:40:45 <andythenorth> if an electric motor can deliver 100HP at stall
12:40:53 <Eddi|zuHause> i'm not a physicist
12:40:58 <andythenorth> why does the gearing affect acceleration?
12:41:59 <andythenorth> 2:1 will produce twice the torque of 1:1, but the RPMs will be half
12:42:10 <andythenorth> so distance travelled over time should be equal in both cases
12:43:45 <Eddi|zuHause> iirc it moves the point where power/speed takes over from max TE
12:43:48 <andythenorth> web pages I'm reading say "this isn't intuitive, unless you're newton" :P
12:43:55 <Eddi|zuHause> but i don't remember in which direction
12:44:30 <Eddi|zuHause> but really, i'm not the person to discuss this. if you find one, please show me the results :p
12:45:48 <Eddi|zuHause> i think with diesel engines, you want to reach a minimum RPM as quick as possible
12:46:11 <andythenorth> you want to keep the RPM in the efficient power band, which can be very narrow on large diesels
12:46:32 <andythenorth> it varies a bit by transmission type (hydraulic, mechanical, electrical)
12:46:50 <Eddi|zuHause> with higher gearing, you can reach that RPM quicker, but you lose max speed from the max RPM
12:46:58 <andythenorth> if it's diesel electric, the key factor is rotational speed of the traction motor
12:47:35 <andythenorth> hmm, so IRL, you can put max stall torque through any electric motor, but not for long
12:48:02 <Eddi|zuHause> there's an I^2 t thing with electric motors
12:48:16 <andythenorth> magic smoke escapes
12:48:54 <Eddi|zuHause> i have no real clue what that does, but it regulates how much current can flow in which timespan
12:49:20 <Eddi|zuHause> which in my limited experience means "if the motor didn't move for 5s, it cuts out"
12:49:23 <andythenorth> there are factors like thermal performance, degradation of the windings
12:49:27 <andythenorth> and risk of flashover
12:49:38 <andythenorth> flashover tends to lead to late arrival of vehicle at destination :P
12:49:56 <andythenorth> so lower gearing means you can put max torque through the traction motor for longer
12:50:28 <andythenorth> somewhere in ground_vehicle I want to add 'torque factor' and default it to 1 :P
12:51:12 <peter1138> Hmm, wonder if I can make jenkins do my deployment on tags.
12:51:35 <andythenorth> green light deploy? :P
12:53:17 <peter1138> I don't deploy directly, but it'd be nice to be able to use the build result in jenkins rather than just building locally.
12:59:18 <andythenorth> force = force * torque_factor :P
13:12:00 <peter1138> Is it a factor or a maximum?
13:12:21 <peter1138> If it's a factor, you might as well just specify a lower power.
13:14:08 <andythenorth> if the power is too low, they don't appear to reach max speed, but yeah
13:14:41 <andythenorth> don't we have to have enough power to overcome resistance?
13:15:03 <peter1138> That's what the kickoff acceleration is about.
13:15:36 <peter1138> Well, that actually means, "err, speed is zero, we can't divide by 0, so just apply max welly"
13:23:02 <DorpsGek_II> [OpenTTD/OpenTTD] fsimonis commented on pull request #6817: PR for Modernizing for C++11: replacing instances of SmallVector https://git.io/fhaQC
13:23:56 <DorpsGek_II> [OpenTTD/OpenTTD] fsimonis commented on pull request #6817: PR for Modernizing for C++11: replacing instances of SmallVector https://git.io/fhaQu
13:26:01 <DorpsGek_II> [OpenTTD/OpenTTD] fsimonis commented on pull request #6817: PR for Modernizing for C++11: replacing instances of SmallVector https://git.io/fhaQD
13:28:18 <DorpsGek_II> [OpenTTD/OpenTTD] fsimonis commented on pull request #6817: PR for Modernizing for C++11: replacing instances of SmallVector https://git.io/fhaQF
13:28:19 *** snail_UES_ has joined #openttd
13:28:30 <DorpsGek_II> [OpenTTD/OpenTTD] fsimonis commented on pull request #6817: PR for Modernizing for C++11: replacing instances of SmallVector https://git.io/fhaQb
13:29:41 <DorpsGek_II> [OpenTTD/OpenTTD] fsimonis commented on pull request #6817: PR for Modernizing for C++11: replacing instances of SmallVector https://git.io/fha7v
13:33:41 <DorpsGek_II> [OpenTTD/OpenTTD] fsimonis commented on pull request #6817: PR for Modernizing for C++11: replacing instances of SmallVector https://git.io/fha7Y
13:41:55 <DorpsGek_II> [OpenTTD/OpenTTD] fsimonis commented on pull request #6817: PR for Modernizing for C++11: replacing instances of SmallVector https://git.io/fha71
13:43:11 <DorpsGek_II> [OpenTTD/OpenTTD] fsimonis commented on pull request #6817: PR for Modernizing for C++11: replacing instances of SmallVector https://git.io/fha7h
13:53:17 <DorpsGek_II> [OpenTTD/OpenTTD] fsimonis commented on pull request #6817: PR for Modernizing for C++11: replacing instances of SmallVector https://git.io/fha5X
13:55:04 <DorpsGek_II> [OpenTTD/OpenTTD] fsimonis commented on pull request #6817: PR for Modernizing for C++11: replacing instances of SmallVector https://git.io/fha5d
13:58:31 <Eddi|zuHause> andythenorth: i think what you're looking for is a more dynamic calculation of "power", instead of assuming a constant power over all circumstances
13:58:50 <andythenorth> I wonder if it's a torque curve
13:59:05 <DorpsGek_II> [OpenTTD/OpenTTD] Eddi-z dismissed a review for pull request #7000: Some NewGRF variables concerning railtypes https://git.io/fhu7c
13:59:14 <andythenorth> literally putting numbers through L160 of ground_vehicle.cpp to see how it works
14:01:28 <DorpsGek_II> [OpenTTD/OpenTTD] Eddi-z commented on pull request #7000: Some NewGRF variables concerning railtypes https://git.io/fhadg
14:02:15 <DorpsGek_II> [OpenTTD/OpenTTD] fsimonis commented on pull request #6817: PR for Modernizing for C++11: replacing instances of SmallVector https://git.io/fhadV
14:03:53 <Eddi|zuHause> andythenorth: there is two main things to consider: 1) how the power output changes with internal RPM, and 2) how the power output changes by runtime
14:05:08 <andythenorth> I wonder if speed is enough of a proxy
14:05:36 <andythenorth> currently all vehicles will develop max power at low speed
14:05:55 <andythenorth> but depending on transmission characteristics, that's not the case for all RL vehicles
14:09:54 <peter1138> It's okay for electrics, but not diesels. I have no idea about steam.
14:11:52 <Eddi|zuHause> CI not working again?
14:12:24 <peter1138> TrueBrain, probably hasn't logged in recently enough :/
14:12:29 <peter1138> That requirement is fecking dumb.
14:12:38 <DorpsGek_II> [OpenTTD/OpenTTD] fsimonis commented on pull request #6817: PR for Modernizing for C++11: replacing instances of SmallVector https://git.io/fhaFC
14:12:50 <Eddi|zuHause> we really need a bot for that
14:12:59 <andythenorth> mechanical turk :
14:13:22 <Eddi|zuHause> one of those birds that flip up and down
14:13:30 <Eddi|zuHause> like in the simpsons
14:16:21 <DorpsGek_II> [OpenTTD/OpenTTD] fsimonis commented on pull request #6817: PR for Modernizing for C++11: replacing instances of SmallVector https://git.io/fhaFK
14:17:05 <DorpsGek_II> [OpenTTD/OpenTTD] fsimonis commented on pull request #6817: PR for Modernizing for C++11: replacing instances of SmallVector https://git.io/fhaFD
14:17:25 <andythenorth> so...a parametric equation for torque vs speed? :P
14:20:10 <Eddi|zuHause> andythenorth: can you check whether #7000 still works for you? (remember new var number and bitshift)
14:20:25 <andythenorth> yes, but not now :)
14:20:28 <andythenorth> I broke...everything
14:20:57 <peter1138> Putting "rm -rf $PATH" in a Makefile feels dirty.
14:21:21 <andythenorth> what could go wrong?
14:21:23 <Eddi|zuHause> peter1138: "what's the worst thing that can happen"?
14:21:32 <andythenorth> I've seen a video of recursive rm -r
14:22:16 <peter1138> Needs -r because directories, needs -f because it's removing a git clone which by default has files which are not writable.
14:22:25 <Eddi|zuHause> I HERD U LIEK RECURSION SO WE PUT SOME RECURSION IN UR RECURSION?
14:22:28 <DorpsGek_II> [OpenTTD/OpenTTD] fsimonis commented on pull request #6817: PR for Modernizing for C++11: replacing instances of SmallVector https://git.io/fhabn
14:22:36 <peter1138> $PATH comes from mktemp -d
14:23:35 <peter1138> Basically this is: make temp directory, clone release repo into it, delete files, copy current files, add to git, commit to git, make tag, push to git, remove temp directory
14:23:50 <peter1138> andythenorth, again, yes. I already ate, but I have some food left over.
14:23:57 <Eddi|zuHause> there was a case a few years ago where a broken "make install" instead of "rm -rf /path" did "rm -rf / path"
14:24:09 <peter1138> Eddi|zuHause, that's the worry.
14:24:24 <andythenorth> I had pork and rice, now I might have mac and cheese
14:24:28 <andythenorth> just snacks really
14:24:41 <peter1138> I had, um, salad, and a cheese salad roll, and some fruit cake.
14:25:13 <peter1138> I have left 3 items of fruit, and 4 various "bars" of minimal goodness.
14:25:17 <andythenorth> oof I let unit tests use all my CPU cores
14:25:20 <andythenorth> very noisy fan now
14:25:30 <andythenorth> can hear it even with headphones on
14:25:38 <peter1138> Worst is the white chocolate Lion bar.
14:26:03 <andythenorth> you have Big Clive level of snacks
14:26:22 <peter1138> I bring them in with my lunch but have stopped eating all of them with lunch. So I have a pile.
14:26:55 <andythenorth> in my office, other people would eat them
14:27:05 <peter1138> They're in my personal drawer, so fuck them.
14:27:17 <peter1138> There is a blueberry muffin the kitchen, free to anyone who dares first.
14:27:28 <peter1138> That's worse than anything in my pile.
14:29:25 <andythenorth> both trains left depot at same time
14:30:05 <andythenorth> whatever the reality is
14:30:16 <andythenorth> I want to nerf train 7 a bit, without stopping it reaching max speed
14:30:25 <andythenorth> I don't care about the physics, except to understand it
14:30:51 <peter1138> Train stats are hard to find. I bet torque is near impossible.
14:31:57 <andythenorth> actually, rather than nerfing train 7, I want to get train 6 accelerating faster
14:32:40 <andythenorth> if train 7 tries to put all it's HP down, it should slip
14:33:06 <DorpsGek_II> [OpenTTD/OpenTTD] fsimonis commented on pull request #6817: PR for Modernizing for C++11: replacing instances of SmallVector https://git.io/fhaNC
14:33:52 <peter1138> Yeah but max te is quite high.
14:34:26 <peter1138> You could make slopes steeper but I'm assuming you're only using slopes to make the effects more obvious.
14:34:59 <andythenorth> I tested longer trains to see the effect of max TE, and it really only shows when both trains are on their knees
14:37:38 <peter1138> It's limiting force to max_te, when accelerating.
14:38:18 <andythenorth> so we can't exceed x kN force
14:38:18 <peter1138> That's pretty much all max_te does.
14:39:13 <andythenorth> ok so train 7 is actually pretty close to being capped by TE
14:39:15 <peter1138> Maybe a possible issue? max_te isn't affected by slopes.
14:39:31 <andythenorth> @calc (3200 * 18) / (50 * 5)
14:39:44 <andythenorth> max TE on that engine is 235kN
14:39:54 <andythenorth> is force in N or kN?
14:40:32 <peter1138> So resistance goes up because of the slope
14:41:44 <peter1138> Means max_te is reached with less speed.
14:42:05 <andythenorth> ok max_te is in N apparently
14:42:29 <peter1138> That's what it says :p
14:43:22 <andythenorth> I was reading L61 too :P
14:44:13 <peter1138> That's the one that was 1000 * 10 until fairly recently.
14:46:11 <peter1138> return _settings_game.vehicle.train_slope_steepness;
14:46:18 <peter1138> Oh! That already is a setting. Nice.
14:46:31 <andythenorth> same for RVs too
14:53:24 <andythenorth> @calc (3000 * 18) / (50 * 5)
14:53:46 <peter1138> I think I'll have a kiwi fruit for a snack later.
14:54:32 <peter1138> I can't remember how accurate *18/5 is :p
14:54:36 <andythenorth> I think I'll try adjusting "speed * 5" to see if interesting results happen
14:55:05 <andythenorth> it's not supposed to be Trainz, right? :P
14:56:22 <peter1138> It's not a magic equation, it's just a conversion done integerly.
15:12:53 <andythenorth> I changed the 5 to 1
15:13:04 <andythenorth> got the desired result
15:13:14 <andythenorth> but it's very 'magic numbers' and specific to 2 trains
15:13:23 <andythenorth> and untested for slope steepness blah blah
15:14:09 <andythenorth> the 65mph train gets to 65mph faster than the 105mph train
15:16:08 <andythenorth> given enough tiles, the 105mph train then passes it
15:16:40 <andythenorth> if they're both capped to 65mph, the 105mph train is worse, even though it has more HP, it loses due to lower TE
15:17:12 <andythenorth> afaict, the main effect is moving the point where force > max TE
15:17:34 <andythenorth> with bonus unknown side effects :(
15:19:15 *** WWacko1976-work has joined #openttd
15:30:21 <Samu> - Use the Dice.png provived by Supermop reflecting a random ai script to replace the newspaper icon.
15:30:44 <Samu> i still dont know how to do this
15:33:12 *** andythenorth has joined #openttd
15:34:15 <supermop_work> hello andythenorth
15:38:51 <supermop_work> hst looks good Pikka
15:42:34 <DorpsGek_II> [OpenTTD/OpenTTD] nikolas updated pull request #7086: Change #6173: Update SDL driver to use SDL 2.0 https://git.io/fhamZ
15:44:08 <Samu> Flyspray clean up, so closing thanks. If this offends, discuss in irc, thanks.
15:47:04 <peter1138> andythenorth, increase the slope factor instead. Cos *18/5 is based on physics ;p
15:52:58 <andythenorth> slope just nerfs them both :)
15:53:09 <andythenorth> the one with less TE eventually gives up
15:53:15 <andythenorth> before the one with more TE
15:53:42 <andythenorth> but it's 1mph vs about 13mph
15:56:02 <Samu> how do you title Reverses in commit message?
15:56:27 <Samu> wanna avoid commit checker failure
15:57:24 <milek7> Add|Feature|Change|Remove|Codechange|Cleanup|Fix|Revert|Doc|Update|Prepare
15:57:49 <andythenorth> you shouldn't need to have a revert commit, and you can get in a mess doing them
15:58:16 * andythenorth speaks with experience :P
15:58:26 <Samu> Remove seems more fitting
15:58:36 <Samu> it wasn't something that was in master
16:01:29 <milek7> you probably should squash it
16:02:41 <Samu> I'm now removing code, unpatching
16:03:45 <Samu> you know, patching in reverse
16:05:21 <milek7> so you should squash it, adding and removing code back and forth in single PR isn't useful
16:07:23 <Samu> the PR was 1 single commit
16:08:46 <Samu> yes, because I'm removing stuff from the initial one
16:09:28 <milek7> so why you don't squash last commit into first?
16:10:03 <Samu> I thought I couldn't do that
16:10:41 <milek7> you can change order here and 'fixup' into previous
16:10:47 <Samu> hmm how am i squashing the 3rd onto first?
16:11:00 <milek7> 'These lines can be re-ordered; they are executed from top to bottom.'
16:11:23 <Samu> ok, let's see what i can do, brb
16:16:41 *** Wormnest has joined #openttd
16:18:47 <Samu> how do I write strikethrough text
16:29:09 <LordAro> are you expecting an answer, or are you unable to read andy's last comment?
16:29:43 <LordAro> learn to use git bisect and find out yourself :p
16:37:53 *** raphikov has joined #openttd
16:43:02 <Samu> andythenorth, I found your answer
16:43:15 <andythenorth> thanks for highlighting me to mention that
16:43:43 *** Thedarkb-T60 has joined #openttd
16:44:30 <LordAro> Samu: a comment on the issue itself would've sufficed
16:46:44 <Eddi|zuHause> there are definitely people who are on the wrong side of the von Moltke spectrum
16:47:08 *** raphikov has joined #openttd
16:50:26 <andythenorth> well that's knocked me out of flow :P
16:50:34 * andythenorth was deep in enterprise software :P
16:55:18 <Eddi|zuHause> video doesn't play
16:55:40 <Eddi|zuHause> no compatible format found, or somesuch
16:56:26 <milek7> ..but what's purpose of that?
16:56:54 <planetmaker> what does it show? :)
16:57:45 <milek7> chain reaction of mouse traps?
16:59:06 <Eddi|zuHause> yeah, i could also run it in vlc, but in firefox it's a "nope"
16:59:22 <Samu> interesting, r27787 is already fixed on my PR, but how come
17:00:07 <andythenorth> what does 'it is already fixed on my PR' mean? :)
17:00:16 <Eddi|zuHause> "Install libavcodec56, libavcodec57, libavcodec58, libavformat56, libavformat57, libavformat58, libavdevice56, libavdevice57, libavdevice58 package from Packman or VLC repository."
17:00:20 <Eddi|zuHause> i might try that
17:00:27 <Samu> it means, I didn't do the fix, but it was carried over to it
17:00:37 <andythenorth> you rebased, no?
17:00:37 <Samu> must have been the rebase
17:01:42 <Eddi|zuHause> hm, i have those already from packman
17:01:58 <Samu> considering the amount of changes i made on ai_gui.cpp, it's surprising
17:02:24 <andythenorth> clearly time for mac + cheese
17:02:30 <andythenorth> and running the tests to see what I broke
17:02:43 <Eddi|zuHause> you think that might fix your keyboard?
17:03:11 <andythenorth> to fix the keyboard I need to pop some keycaps off, mostly the t key this week
17:03:18 <andythenorth> also I need to type better
17:03:25 <Eddi|zuHause> i wouldn't think that cheese would help :p
17:03:42 <andythenorth> 11 tests broken so far
17:04:05 <andythenorth> what could be better? :P
17:06:54 <Samu> TODO: remove all windows closures to test them again and see if they still crash
17:08:02 <Samu> ... because they could have been fixed already without my knowledge
17:09:04 <Eddi|zuHause> installing more stuff didn't help :/
17:09:51 <Eddi|zuHause> maybe i need to restart some stuff
17:09:54 <andythenorth> how about....acceleration callback? o_O
17:10:01 <andythenorth> and I do a custom calculation in newgrf?
17:10:01 <Eddi|zuHause> andythenorth: no
17:10:10 <andythenorth> eh is TE in cb36? o_O
17:10:19 <andythenorth> and on what triggers?
17:10:29 <Eddi|zuHause> andythenorth: if it is, it won't be recalculated often enough
17:10:46 <andythenorth> there's probably a feature request about that
17:10:52 <Eddi|zuHause> andythenorth: mostly on station load and railtype change
17:11:35 <andythenorth> it's probably needed for the USSR set, right?
17:12:00 <Eddi|zuHause> andythenorth: using any other condition might be a desync hazard
17:12:15 <andythenorth> ok we'll file that one
17:12:32 <Eddi|zuHause> andythenorth: and calling more often will be a performance hazard
17:12:45 <andythenorth> would be 'interesting' though
17:13:21 <Eddi|zuHause> more like "nightmare"
17:18:28 *** synchris has joined #openttd
17:19:31 <DorpsGek_II> [OpenTTD/OpenTTD] nielsmh commented on pull request #7086: Change #6173: Update SDL driver to use SDL 2.0 https://git.io/fhV8W
17:21:03 <nielsm> andythenorth, well if you're claiming to be a simulator of driving trains, you damn better also simulate things correctly
17:21:49 <andythenorth> that's my new train :P
17:22:37 <nielsm> I stopped playing railworks/TS after continually getting fed up with bugs and bad simulation of various things, and heard bad things about TSW as well so haven't tried that at all
17:28:37 <peter1138> Urgh. You follow a guideline for doing something with Jenkins and... none of it is right :/
17:31:57 <Eddi|zuHause> andythenorth: i'm just reading a bunch of "this is fake, and the earth is flat."
17:40:52 <DorpsGek_II> [OpenTTD/OpenTTD] SamuXarick commented on issue #6453: Crash: scrolling down list of (none) GS parameters https://git.io/fhVB9
17:44:29 <Samu> 2019-01-23T16:37:32.1396664Z ##[warning]src\lang\latvian.txt(3904,0): Warning : String name 'STR_AI_CONFIG_HUMAN_PLAYER' does not exist in master file (warning)
17:44:36 <Samu> i dont get this warning here
17:44:57 <nielsm> you removed a string from english.txt
17:45:06 <nielsm> but did not remove it from all other languages
17:45:13 <Samu> i did, only latvian is complaining
17:46:37 <DorpsGek_II> [OpenTTD/OpenTTD] nikolas commented on pull request #7086: Change #6173: Update SDL driver to use SDL 2.0 https://git.io/fhVRV
17:46:39 <TrueBrain> <@peter1138> TrueBrain, will there be Linux binaries for nightlies? <- no clue; someone has to make the Dockerfiles for it. Till then, no.
17:46:47 <nielsm> there's no changes to latvian.txt in your branch
17:47:05 <Samu> but how come, I cleared it from all of them
17:50:45 *** Thedarkb-T60 has joined #openttd
17:51:34 <Samu> I don't have that line on my latvian.txt
17:52:05 <Samu> ah, found it, nevermind, I'm probably drunk
17:52:17 <Samu> STR_AI_CONFIG_HUMAN_PLAYER :Spēlētājs (cilvēks)
17:58:46 <supermop_work> andythenorth: i feel like those comments sum up all of my experience with uk enthusiasm
18:00:16 <supermop_work> kicking and screaming that near 50 year old trains might get displaced from mainline service
18:01:45 <supermop_work> probably a bit just under the surface about the perceived nationality of the purveyor
18:02:01 *** andythenorth is now known as Guest1406
18:02:01 *** andythenorth has joined #openttd
18:03:08 *** Progman has joined #openttd
18:07:47 <andythenorth> OTOH, I was sad that the HST would be gone on GW main line
18:08:03 *** Gustavo6046 has joined #openttd
18:08:04 <andythenorth> but I went on the new train last week, it's fricking awesome
18:08:22 <andythenorth> only complaint is the interior lighting design
18:08:32 <andythenorth> super-white LED strips
18:12:54 <supermop_work> should dreamliner it
18:13:10 <supermop_work> adjust temperature throughout day
18:13:35 <supermop_work> hst is my favorite train but
18:15:19 <andythenorth> seats are spartan, but not bad
18:15:21 <andythenorth> but those lights
18:15:28 <andythenorth> at 06.15 in winter, not wanted
18:17:49 <nielsm> do the seats rotate 180 deg like in shinkansen?
18:25:20 <nielsm> I guess the personnel required to flip all seats at each end of the line would be too costly
18:25:35 <nielsm> because it's awesome for comfort
18:30:31 *** HerzogDeXtEr has joined #openttd
18:34:05 *** sla_ro|master has joined #openttd
18:37:48 <Borg> it was legendary PikkaBird?
18:45:40 *** supermop_work_ has joined #openttd
18:45:56 *** HerzogDeXtEr1 has joined #openttd
18:46:33 *** supermop_work__ has joined #openttd
18:53:45 *** supermop_work_ has quit IRC
19:01:52 *** frosch123 has joined #openttd
19:09:00 <Samu> which window is QUERY_STRING
19:09:36 <Samu> is it the one where we manually enter values? cancel/ok?
19:10:36 <Samu> ok, in that case, the svr r27787 fix is incomplete, should also close the QUERY_STRING window
19:10:58 <Samu> because when an AI starts, the start_date parameter is no longer editable
19:12:05 <Samu> or maybe, only some query_string windowses should be closed
19:12:12 <Samu> those that are no longer editable
19:13:23 <Samu> Utf8Decode(WChar *c, const char *s) still crashes
19:13:59 *** Thedarkb-X40 has joined #openttd
19:15:08 <Samu> DeleteWindowByClass(WC_TEXTFILE); // TODO: this is needed to avoid openttd crash, unless there's another way to manage this window
19:15:29 <Samu> crash still present on master
19:16:18 <Samu> Wtf8Decode complains about null pointers
19:17:34 <Samu> i think it's because the text file is switched to nothingness
19:17:47 <Samu> then it tries to decode the nothing and complains about null
19:18:07 <glx> you can check the call stack
19:18:18 <glx> and see exactly what happens
19:18:19 <Samu> if i have the license.txt window open
19:18:35 <Samu> and an AI is switched to another
19:19:05 <Samu> also happens with changelog.txt and readme.txt
19:19:29 <Samu> and if it wasn't fixed here, i guess it is also present on newgrf's similar .txt files
19:20:54 <Samu> my fix is "close before it crashes"
19:24:31 <Samu> Exception thrown: read access violation.
19:26:49 <DorpsGek_II> [OpenTTD/OpenTTD] glx22 commented on pull request #7077: Change: Always build baseset metadata, instead of including it in the repo. https://git.io/fhVwn
19:29:08 <glx> was talking about DorpsGek_II
19:29:26 <Samu> + s 0x0000000000000000 <NULL> const char *
19:29:39 <Samu> the current value of s at that point
19:30:11 <glx> now walk through the call stack to see why
19:32:33 <DorpsGek_II> [OpenTTD/OpenTTD] glx22 updated pull request #7022: Add: generate_widget.vbs to allow script_window.hpp enums generation … https://git.io/fhsWa
19:33:26 <DorpsGek_II> [OpenTTD/OpenTTD] nielsmh approved pull request #7022: Add: generate_widget.vbs to allow script_window.hpp enums generation … https://git.io/fhVwD
19:35:15 <glx> you can double click on call stack lines
19:37:40 <Samu> + buffer 0x00000023ab6fe970 "î†AI licence of \x4" char[2048]
19:38:52 <glx> some part of an utf8 encoded character
19:39:36 <glx> you just get a preview of the buffer
19:40:22 <glx> as it has an address and some content
19:41:40 <Samu> the crash is easy to reproduce
19:42:16 <Samu> without closing changelog, select another AI, pick AdmiralAI which has no changelog
19:42:22 <Samu> and it crashes right away
19:42:39 <LordAro> be sure to include which AI versions in your bug report :p
19:43:14 <Samu> i opened it some years ago :( it's lost somewhere
19:43:53 <glx> all issues from previous bug tracker are on github
19:44:46 <DorpsGek_II> [OpenTTD/OpenTTD] glx22 merged pull request #7022: Add: generate_widget.vbs to allow script_window.hpp enums generation … https://git.io/fhsWa
19:46:30 <Samu> i better make a new issue
19:46:33 <Samu> that one is getting confusing
19:46:39 <Samu> part of it is fixed, some other part is not
19:46:47 *** gelignite has joined #openttd
19:47:39 <LordAro> Samu: yes, looks like aeveral issues in one there, best make a new one
19:52:53 <DorpsGek_II> [OpenTTD/OpenTTD] SamuXarick opened issue #7088: Crash when changelog.txt, readme.txt, license.txt do not exist https://git.io/fhVoc
19:56:06 <Samu> the issue reported about newgrf is still valid
20:01:08 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN commented on pull request #7077: Change: Always build baseset metadata, instead of including it in the repo. https://git.io/fhVob
20:01:50 *** supermop_work has joined #openttd
20:02:12 <DorpsGek_II> [OpenTTD/OpenTTD] nielsmh commented on pull request #7077: Change: Always build baseset metadata, instead of including it in the repo. https://git.io/fhVKv
20:03:24 *** supermop_work_ has joined #openttd
20:03:55 *** supermop_work__ has quit IRC
20:09:54 <Samu> hmm the newgrf crash is a little bit different
20:09:57 <Samu> const char *name = GetGRFStringFromGRFText(this->name->text);
20:10:24 <Samu> Exception thrown: read access violation.
20:14:06 <Samu> do i make a new issue for newgrf?
20:15:06 <Samu> name = 0xcccccccccccccccc <Error reading characters of string.>
20:16:47 <Samu> i guess i'll do it anyway
20:31:26 <LordAro> Samu: i cannot reproduce any crash for #7088
20:31:39 <LordAro> the title of the changelog window changes, but the text stays the same and it doesn't crash
20:40:01 <Samu> well, i'm not getting a crash
20:40:17 <Samu> maybe this build is still mine?
20:41:29 <Samu> sorry, looks like I was testing the wrong build
20:42:33 <Samu> no crash with my build now
20:43:17 <Samu> are you testing in debug mode?
20:43:23 <Samu> gonna try in release mode
20:44:01 <DorpsGek_II> [OpenTTD/OpenTTD] andythenorth commented on issue #7089: Crash when Removing NewGRF from Active list https://git.io/fhVP8
20:44:06 <LordAro> i'm going to go ahead and close the issue, anyway
20:44:17 <Samu> i swear im getting a crash somehow
20:44:26 <Samu> maybe only in release mode, let me confirm
20:44:35 <LordAro> if you can reproduce it, remember to upload the crash.* files
20:45:36 <Samu> I can't get this to crash now... why :|
20:45:51 <Samu> I swear it crashed an hour ago, by doing those exact steps
20:46:00 <LordAro> sounds like it wasn't actually a master build :p
20:46:09 <planetmaker> typical heisen-bug ;)
20:47:35 <Samu> switch to Random AI from AIAI
20:56:04 <DorpsGek_II> [OpenTTD/OpenTTD] SamuXarick commented on issue #7088: Crash when changelog.txt, readme.txt, license.txt do not exist https://git.io/fhVXS
21:11:08 <glx> btw dmp is only valid for build from the compile farm
21:32:25 <TrueBrain> yippie, I can now trigger builds from releases; means the website now auto-updates when ever there was a release :)
21:32:55 <TrueBrain> if building it wouldnt fail :D
21:38:58 <DorpsGek_II> [OpenTTD/binaries-proxy] TrueBrain opened pull request #6: Add: url for health probe, which always returns 200 https://git.io/fhVyw
21:41:53 <DorpsGek_II> [OpenTTD/website] TrueBrain opened pull request #34: Add: url for health probe, which always returns 200 https://git.io/fhVyA
21:43:35 <DorpsGek_II> [OpenTTD/binaries-proxy] TrueBrain merged pull request #6: Add: url for health probe, which always returns 200 https://git.io/fhVyw
21:49:32 <LordAro> so... i can't post comments
21:50:16 <LordAro> "You can't comment at this time."
21:53:20 <DorpsGek_II> [OpenTTD/OpenTTD] TrueBrain commented on issue #7088: Crash when changelog.txt, readme.txt, license.txt do not exist https://git.io/fhVSX
21:53:38 <TrueBrain> so sadly, it is a local issue :(
21:54:23 <DorpsGek_II> [OpenTTD/OpenTTD] LordAro commented on issue #7088: Crash when changelog.txt, readme.txt, license.txt do not exist https://git.io/fhVSQ
21:55:18 <DorpsGek_II> [OpenTTD/website] TrueBrain merged pull request #34: Add: url for health probe, which always returns 200 https://git.io/fhVyA
21:55:47 <TrueBrain> and I fixed the proxy :D
21:56:58 <TrueBrain> only thing I have to think about how to produce a new version of the production container .. as strictly seen you first want that to go to staging
21:57:09 <TrueBrain> but ... doing that every night is .. euh ... not really cool
21:59:49 <TrueBrain> okay, so I am just going to rebuild production image and roll it out .. but .. how do I know latest tag for production .. hmm
21:59:55 <TrueBrain> not for this night I guess :)
22:01:00 <TrueBrain> tomorrow it will be available around 20:15 CE(S)T :D
22:03:21 <andythenorth> supermop_work_: so narrow gauge steel mill trains?
22:11:26 <Samu> well, slot -1 is the slot for Random AI
22:12:04 <Samu> it doesn't exist and exists at the same time
22:17:46 <Samu> oh wait nvm, wrong window
22:21:08 <Samu> hmm so... random AI needs a name
22:26:18 *** Thedarkb-T60 has joined #openttd
22:27:56 <Samu> i have an idea for the name
22:28:46 <Samu> Change(NULL) is fine, it's changing to random ai
22:36:00 <supermop_work_> andythenorth: yes
22:36:27 <andythenorth> industrial trains :P
22:36:38 * andythenorth has whole new idea :P
22:44:08 <andythenorth> industrial trains don't fit in Horse right now
22:44:19 <andythenorth> because they need to be slower, for larger capacity
22:51:06 <Eddi|zuHause> unfortunately i cannot close tickets
22:51:51 <Eddi|zuHause> andythenorth: i'm pretty sure that ticket can be closed as invalid
22:52:39 <DorpsGek_II> [OpenTTD/OpenTTD] andythenorth commented on issue #7087: Presignal doesn't work on long distance https://git.io/fhV51
23:07:00 <Samu> what should it be doing?
23:25:10 <Samu> error C2440: '=': cannot convert from 'char *' to 'const char'
23:28:21 <Samu> i went with 1024, but not sure if that's appropriate
23:46:36 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN commented on issue #7062: Ship pathfinders ignore the max order distance https://git.io/fhVAt
23:48:02 <DorpsGek_II> [OpenTTD/OpenTTD] nikolas updated pull request #7086: Change #6173: Update SDL driver to use SDL 2.0 https://git.io/fhamZ
continue to next day ⏵