IRC logs for #openttd on OFTC at 2018-05-19
⏴ go to previous day
00:09:22 *** Smedles has joined #openttd
00:14:57 <peter1138> Hmm... how to store a value which can be a DiagDir or an Axis?
00:36:27 *** Thedarkb has joined #openttd
00:40:57 <Eddi|zuHause> have it be a DiagDir and for the cases where it's an Axis assert it doesn't use any values outside DiagDirToAxis()?
00:41:50 <Eddi|zuHause> or use modern concepts like polymorphism
00:42:27 <peter1138> That's not so helpful for stuffing into 3 bits in the map array.
00:42:55 <peter1138> Nor do I want to rewrite the whole game.
00:43:07 <Eddi|zuHause> but as far as i know, Axis is a subset of DiagDir, right?
00:43:34 <peter1138> DiagDir is compass directions.
00:44:19 <Eddi|zuHause> yeah, you have 2 axes and 8 diagdirs
00:45:37 <Eddi|zuHause> ah, 8 directions, 4 diagdirs and 2 axes
00:46:14 <peter1138> Right, but that's irrelevant.
00:46:24 <Eddi|zuHause> but there definitely were these conversion functions like "AxisToDiagDir"
00:46:54 <peter1138> No, DiagDir or Axis.
00:48:22 <Eddi|zuHause> why is there need to differentiate that? surely the axis is just the diagdir modulo up/down
00:48:54 <Eddi|zuHause> NE/SW == X, NW/SE == Y, or so
00:49:13 <peter1138> DiagDir is half a tile (single direction), Axis is a full tile (bi directional)
00:49:49 <Eddi|zuHause> i'm afraid i'm missing some context on what you're trying to achieve
02:51:30 *** rocky1138 has joined #openttd
03:00:13 *** Thedarkb1 has joined #openttd
03:01:12 *** sim-al2 has joined #openttd
04:28:20 *** muffindrake1 has joined #openttd
05:52:52 *** sim-al2 is now known as Guest3074
05:52:54 *** sim-al2 has joined #openttd
06:04:03 *** sla_ro|master has joined #openttd
08:11:34 *** andythenorth has joined #openttd
08:40:22 *** Alberth has joined #openttd
08:40:22 *** ChanServ sets mode: +o Alberth
08:55:48 *** andythenorth has joined #openttd
09:39:50 *** Progman has joined #openttd
09:46:45 <andythenorth> what we've got here is failure to communicate
09:51:10 <peter1138> Communication breakdown
10:02:47 <andythenorth> some men you just can't reach
10:02:57 <andythenorth> I don't like it any more than you do
10:03:40 <Alberth> right, openttd dropping revision number breaks BB
10:06:37 <andythenorth> wonder if it breaks grfs too? o_O
10:07:56 <Alberth> I don't see how it won't break grfs that use anything "introduced in rXYZ"
10:08:32 <andythenorth> maybe we need an adapter
10:08:43 <andythenorth> to preserve old revs :P
10:09:07 <andythenorth> or simply return that the github build is > any old rev number used in GS or newgrf
10:09:11 <Alberth> wouldn't just setting it to the latest rev work?
10:10:15 <nielsm> or maybe even rev of the move + day difference of most recent git commit to day of the move
10:10:39 <peter1138> rev of move + number of commits since :p
10:10:50 <peter1138> (Wrong for branches but okay for master)
10:10:59 <nielsm> but is that easy to calculate?
10:11:18 <andythenorth> count hashes since hash xyz?
10:11:23 <andythenorth> it's easy at compile time no?
10:11:55 <nielsm> well it's (at least) linear time on the number of commits, I think?
10:11:57 <andythenorth> it will be interestingly broken for forks / patchpacks :P
10:12:02 <nielsm> (not well-versed in git data structures)
10:12:04 <andythenorth> but that applies to the svn rev too
10:12:44 <andythenorth> len git rev-list or something
10:12:56 <Alberth> if you want new features in-between releases, you do need some incremental number thing
10:14:07 <andythenorth> we could strictly version the API instead
10:14:15 <andythenorth> so check the API rev, not ottd rev
10:14:52 <andythenorth> kind of bureaucratic :P
10:14:57 <andythenorth> needs manual intervention every time
10:15:07 <nielsm> save game version already does
10:15:13 <Alberth> not unlike save game versions :p
10:15:49 <andythenorth> we could have a CI guard for that, but it would be complex
10:16:04 <andythenorth> if there was a hash of the current API features, stored against a version
10:16:17 <andythenorth> then CI could fail if hash changes without rev bumping
10:16:59 <Alberth> you'd need a separate API description too then
10:17:39 <nielsm> or otherwise change from version-checking going forward to use feature detection
10:18:05 <andythenorth> have a test suite, and hash that :P
10:18:12 <andythenorth> layers in layers in layers :P
10:23:23 <peter1138> Hmm, why is forbidding 90 deg turns for ships bad?
10:24:34 <Alberth> lots of narrow passages for ships
10:24:49 <Alberth> anything river or canal eg
10:25:39 <Alberth> even 3 wide isn't enough if the ship is not at one of the sides
10:28:42 <peter1138> Is there any case where forbidding 90 deg turns for ships is good?
10:30:13 <Alberth> it looks better, but that holds equally for trains, I guess
10:30:45 <Alberth> hmm, 1.5 doc has more parameters for publishing news than my fallback
10:31:02 <Eddi|zuHause> <peter1138> rev of move + number of commits since :p <-- how would that work for people checking out stuff from before the move? or forking from there?
10:32:22 <peter1138> Er, before the move they have the rev already.
10:32:30 <peter1138> And forking from before the move... well that's stupid.
10:33:03 <Eddi|zuHause> people do stuopid things all the time
10:33:29 <peter1138> Anyway, if it's a fork, it's always gotta be maintained by them separately.
10:34:38 <Alberth> wouldn't it already break by checking out an old revision?
10:34:39 <peter1138> Alberth, hmm, I guess it would be nice to some allow ships to turn around on the spot (rather than instant reversing)
10:34:51 <Alberth> the revision number isn't in the source itself
10:35:26 <Alberth> that sound like a good idea, peter1138
10:35:30 <Eddi|zuHause> anyway, we really need some sort of finegrained counter, either automatic (commits? nightlies?) or manual (like savegame version)
10:36:04 <peter1138> Issue I see is that ships get to a dock, and pathfind continuing forwards, so they do a loop forwards instead of just turning around.
10:36:05 <Alberth> I wonder if BB doesn't claim incorrect compatibility in this case
10:36:22 <Eddi|zuHause> Alberth: iirc, the revision number is in the commit message, and be build script checked that
10:37:12 <Alberth> yes, so git co <some-old.rev> ; ./configure ; make run fails, as the rev number is not in the source
10:37:32 <Wolf01> TL;DR what is the argument of the day?
10:37:59 <Alberth> would be a bit too obvious, wouldn't it? :)
10:38:19 <Wolf01> I could make a script to convert to ascii everything I write if you like it more ;)
10:38:41 <peter1138> I wonder if going out on the bike will make me feel better
10:38:49 <peter1138> (Got a cold coming on :()
10:39:37 <Wolf01> A cup of tea and a ginger biscuit
10:39:42 <andythenorth> peter1138: sometimes ships with 90º forbidden get stuck in a dead end river
10:40:05 <andythenorth> they leave a dock the 'wrong' way, then pathfinder fails to find a route out of the dead end somehow
10:40:14 <Wolf01> The real question is: why is the ship there?
10:40:17 * andythenorth had the lergy Monday/Tuesday, fell asleep at work
10:40:52 <Alberth> aliens dropped the ship there
10:41:22 * Wolf01 installs train mechanic simulator 2017
10:41:38 <Eddi|zuHause> why did the ship cross the road?
11:41:33 <Wolf01> Yeah, I repaired a diesel engine rubber springs, it was really easy!
11:42:17 <Eddi|zuHause> how are springs made of rubber?
11:42:48 <Wolf01> There are steel coil springs and rubber blocks on the suspension system
11:43:00 <Wolf01> At least the game says that
11:43:24 <Eddi|zuHause> i somehow just try to imagine a surgeon-simulator type game :ü
11:47:59 <Wolf01> Mmmh, brakes cost too much, and broken parts sell for a really low price... I made enough money to complete the current order, but what for the next one?
11:48:42 <Wolf01> Oh, going to scrap a locomotive
11:51:55 *** ToBeFree has joined #openttd
12:00:41 *** synchris has joined #openttd
12:16:32 <Eddi|zuHause> Alberth: i know of that, but that's not really what i meant :)
12:17:59 <Eddi|zuHause> Alberth: surgeon simulator is one of those "fun" games where the controls are so ill-defined that you're bound to kill the person anyway
12:18:52 <Eddi|zuHause> just go to youtube and pick a random video about it :p
12:19:26 <Wolf01> I can't understand the signals in this game
12:19:40 <Eddi|zuHause> that's common for signals :p
12:21:20 <Alberth> signals are a bad feature, we should remove them in the next version
12:24:08 <Wolf01> Better, timetable properly the trains so they won't collite
12:24:37 <Eddi|zuHause> i think that game exists :p
12:25:16 <Eddi|zuHause> was thinking of train valley
12:25:25 <Wolf01> Ok, the most boring part of this game are the screws
12:27:30 <nielsm> A-train doesn't have signals, you have to timetable trains to not interfere
12:28:14 <nielsm> (at least based on the one version I've played)
12:28:33 <peter1138> Now that's an old series.
12:30:35 <nielsm> thing I set up last year
12:32:38 <Eddi|zuHause> they're all driving on the wrong side
12:33:06 <nielsm> it's japan, left is right
12:36:16 *** frosch123 has joined #openttd
12:43:39 <frosch123> peter1138: roadstops have two bits which can be either axis or diagdir
12:43:57 <Wolf01> Ow... I can't keep scrapped parts to repair them for later :(
12:45:05 <peter1138> frosch123, yes, but that depends on what the station graphics value is.
12:46:00 <peter1138> I will probably go with 0-3 = diagdirection, 4 = axis_x, 5 = axis_y
12:47:52 <peter1138> If I bother doing anything :)
12:49:36 <frosch123> that sounds equivalent to bit 2 deciding between axis and diagdir
12:58:17 *** gelignite has joined #openttd
13:06:06 <Wolf01> Driving around the map looks like driving a model train where the coupling is attached to the truck, but I don't really bother, I'm used to the sudden 45 degree turns of OTTD :P
13:09:05 <peter1138> frosch123, pretty much, yes.
13:10:12 *** Eddi|zuHause2 has joined #openttd
13:20:37 <Wolf01> Chance to repair 100% => repair failed
13:22:26 <Wolf01> I can't purchase a new part, I was expected to repair it...
13:24:45 *** Thedarkb1 has joined #openttd
13:37:50 <Wolf01> I could purchase it... loosing the bonus
13:43:13 <Wolf01> "there's a problem on the exhaust system, go out and find the engine"... the engine was laying on the side
14:12:01 *** andythenorth has joined #openttd
14:13:33 <Wolf01> Awww damaged truck frame...
14:13:59 <andythenorth> peter1138: you on a bike then? o_O
14:14:05 *** Coobies has joined #openttd
14:28:00 <peter1138> Thinking about it. Got a cold coming on so might go for something gentle.
14:29:09 *** Eddi|zuHause2 is now known as Eddi|zuHause
14:29:12 <peter1138> So it improves hash collisions, but otherwise no performance improvement.
14:32:40 <peter1138> I did wonder if there's something in std:: to use instead.
14:34:35 *** chomwitt has joined #openttd
14:34:56 * andythenorth is in chesham, weather is tops
14:35:05 <andythenorth> near some kind of Chiltern cycleway
14:35:25 <Eddi|zuHause> you mean like a standard set of hash functions?
14:36:03 <Eddi|zuHause> there's occasionally a tradeoff between the number of collision, and the complexity of calculating the hash
14:38:53 <peter1138> Arr, Chesham is not far from me.
14:53:13 <peter1138> if ((2 * Thash_bits) < 32) hash ^= hash >> (min(2 * Thash_bits, 31));
14:53:22 <peter1138> Given that 2 * Thash_bits is < 32
14:53:36 <peter1138> I don't see the point of the min()
14:53:54 <peter1138> Never mind the redundant brackets.
14:55:24 <peter1138> hash ^= hash >> 24;
14:55:28 <peter1138> hash ^= hash >> 12;
14:55:40 <peter1138> Is what boils down to.
14:55:50 <peter1138> hash -= hash >> 17;
14:55:59 <peter1138> Is the replacement.
14:58:38 <Eddi|zuHause> i can't see a huge performance impact of ^ vs -
14:59:33 <peter1138> Yeah, it looks worse currently use to the if and min conditionals.
14:59:54 <peter1138> But Thash_bits is compile-time constant so they should be optimized out.
15:02:08 *** virtualrandomnumber has joined #openttd
15:07:23 *** sim-al2 is now known as Guest3096
15:07:23 *** sim-al2 has joined #openttd
15:11:54 <_dp_> performance comes from hash function quality, not ^ vs -
15:12:51 <_dp_> haven't checked if it's actually better but overall description sounds pretty reasonable
15:13:23 <_dp_> if hash function collides on straight lines it's not a good function
15:16:31 <peter1138> I think the issue here is that the system using the hash is vastly more costly than the hashtable and so its quality doesn't matter. But it's nice to make it better quality.
15:26:09 <virtualrandomnumber> if there's a conditional order jump and you type in the value "306", it is displayed as "307" after clicking OK
15:26:54 <nielsm> without knowing anything about how it actually works, my guess is that it can be a side effect of how speed values are stored and handled internally
15:33:35 <_dp_> yeah, something gets lost when doing ConvertDisplaySpeedToSpeed and back with ConvertSpeedToDisplaySpeed
15:34:34 <Eddi|zuHause> it's basically a rounding error
15:41:10 <peter1138> There'll also be no vehicle with a max-speed of 306 km/h for the same reason, I guess.
15:41:18 <_dp_> 301->302, 303->304, 306-307, 308->309, 311->312, yeah, looks like rounding error
15:52:35 <andythenorth> who makes this stuff anyway?
15:55:33 <peter1138> Oh... villager is opening a door it can't reach. Nice.
16:05:13 <_dp_> ehm... why do condition orders work in mph instead of game internal units?
16:06:06 <Wolf01> I thought mph were the internal units
16:06:36 <Eddi|zuHause> internal units are weird
16:07:21 <_dp_> which is mph * 10 / 16 iirc
16:08:19 <_dp_> other way around, * 16 / 10
16:08:34 <Eddi|zuHause> some people were feeling 1mph=1.6km/h is not accurate enough
16:09:01 <_dp_> so it's about equal kmh but rounds well with mph %)
16:09:28 <Eddi|zuHause> ... i think catering for that made everything worse
16:10:50 <_dp_> well, whatever, speed orders are screwed because they store in mph
16:10:56 <_dp_> so it's actually a domain error
16:11:10 <_dp_> mph can't store half of kmh values even though game itself can
16:12:35 <_dp_> OrderConditionCompare(occ, v->GetDisplayMaxSpeed() * 10 / 16, value)
16:12:49 <_dp_> that's order condition very logic, like wtf xD
16:13:36 <Eddi|zuHause> how does that not desync when people play with different units?
16:13:44 <andythenorth> conditional orders are a bag of spanners
16:13:57 <_dp_> display there is not display speed)
16:14:45 <_dp_> don't ask me why GetDisplayMaxSpeed returns speed in game units xD
16:15:26 <_dp_> * Returns max speed of the engine for display purposes
16:15:30 <_dp_> * @return max speed in km-ish/h
16:16:04 <Eddi|zuHause> for what other purposes are the other speed functions?
16:16:58 <_dp_> Eddi|zuHause, by the looks of it there are no other functions
16:17:38 <Eddi|zuHause> _dp_: i think you're treading down a wtf-fractal there
16:19:10 <andythenorth> there is too much sun on my screen to fix Hog
16:19:21 <andythenorth> I can read irc because it's high contrast :P
16:19:39 <nielsm> make a high contrast graphics set
16:19:47 <nielsm> oh wait is that toyland?
16:19:53 <_dp_> well, I guess logic was that {VELOCITY} accepts kmh-ish/h values, so returning them counts as "display purposes"
16:20:05 <_dp_> and other stuff just piled on top of existing functions
16:21:39 <_dp_> and nobody noticed that it basically does ConvertKmhishSpeedToDisplaySpeed(v->GetDisplayMaxSpeed()) ...
16:27:53 <_dp_> I bet there is even some max speed value that doesn't work with "equals" condition even if you can put it there
16:28:19 <_dp_> because of different rounding for displaying max speed for vehicle and max speed in condition
16:28:39 <Eddi|zuHause> _dp_: shows how "well tested" this feature is
16:36:37 <andythenorth> and nobody understands conditional orders
16:37:01 * andythenorth *is* irrationally biased against conditional orders, sorry :)
16:37:34 <andythenorth> is there a guide somewhere about what they're for?
16:37:34 <_dp_> Eddi|zuHause, It kind of looks like it was specifically designed to avoid that so I may be wrong
16:38:11 <andythenorth> doesn't explain why they're useful though
16:38:33 <andythenorth> the example is stupid
16:38:45 <andythenorth> why piss around making conditional orders for old vehicles?
16:38:48 <Eddi|zuHause> i always felt like it doesn't cover the most obviously useful cases anayway
16:38:56 <_dp_> I know conditional orders can be exploited for some SRNW
16:38:59 <andythenorth> old vehicles are auto-renewed, it's solving a non-issue
16:39:21 <andythenorth> conditional orders only seem to serve the purpose of giving weird people a topic for forum posts
16:39:27 <andythenorth> and bug reports about 'it doesn't work'
16:39:40 <andythenorth> the only case I ever wanted them for, they can't do
16:39:43 <andythenorth> which is partial loading
16:39:50 <andythenorth> 'leave when load is 70%'
16:40:05 <andythenorth> so that a train can collect from multiple places
16:40:14 <andythenorth> but that actually useful case isn't possible
16:41:18 <_dp_> I'd probably use "renew vehicle if it's older than 3 year" but doesn't seem like they can do it either Xd
16:41:53 <andythenorth> they are as far as I can tell, nonsense
16:42:08 <Eddi|zuHause> andythenorth: i think partial loading is something that must be separate from conditional orders
16:42:35 <Eddi|zuHause> andythenorth: conditional orders can only ever be evaluated when loading has already finished
16:42:53 *** johnwhitlow[m] has quit IRC
16:42:58 <andythenorth> seems like we shoudl delete them
16:43:29 <Eddi|zuHause> andythenorth: they still should be able to do the opposite: "wait until at least 70% full"
16:43:34 *** argoneus_ has joined #openttd
16:43:48 <Eddi|zuHause> [but continue to load to 100% if available]
16:45:39 <andythenorth> the examples I've seen turn off 'full load' and loop the train around constantly
16:45:42 <andythenorth> which seems stupid
16:45:57 <andythenorth> totally logical, but stupid
16:46:53 <Eddi|zuHause> yes, it misses an improvement whereby a "gp tp <same station>" order is checked before leaving the station, and no movement takes place, just goes back to loading state
16:48:59 <_dp_> Seems like current orders can be used for "if train is full jump to unload station"
16:49:16 <_dp_> somewhat reasonable for collecting cargo from a several stations
16:50:06 <Eddi|zuHause> _dp_: yes, but you can only collect from several places if each of them doesn't fill the whole train. you have no way of restricting the load below what's already waiting
16:51:54 <_dp_> Eddi|zuHause, I see no reason for collecting from several when one can fill the train
16:51:55 <Eddi|zuHause> there's also no "try to fully load, but wait at most X days" order
16:52:38 <_dp_> problem with collecting from several is that last station is much less likely to have anything collected from it
16:52:51 <Eddi|zuHause> _dp_: well, the equivilent problem exists for unloading
16:52:53 <_dp_> so don't see how that can be useful in an actual game
16:53:49 <_dp_> Eddi|zuHause, yeah, but unloading can't be done with conditional orders. And I tried to think of some use for them)
16:54:01 <_dp_> Eddi|zuHause, unloading is actualy much more useful with stuff like firs
16:54:54 <Eddi|zuHause> _dp_: well, conditional orderns with unloading works for things like PBI, where the industry stops accepting
16:54:57 <_dp_> Eddi|zuHause, can't you make train wait X days with timetables?
16:55:17 <Eddi|zuHause> _dp_: as in "if load > 0% after unloadingstation, go to overflowstation"
16:55:53 <Eddi|zuHause> _dp_: yes, but you cannot say which one takes precedence, the timetable or the full load order
16:56:07 <_dp_> Eddi|zuHause, don't make it full load?
16:56:30 <Eddi|zuHause> _dp_: you can say "full load or wait for X days", but it fills util full, and THEN waits another few days, if the time has not pased
16:56:41 <Eddi|zuHause> _dp_: you cannot cancel full load if the time is exceeded
16:57:17 <Eddi|zuHause> _dp_: and you can also not make it leave before the time exceeded, if the full load is reached
16:57:50 <_dp_> Eddi|zuHause, I mean just make it load if available and wait X days
16:57:57 <andythenorth> conditional unloading would also be useful
16:57:59 <_dp_> Eddi|zuHause, or does it not load then even if cargo comes?
16:58:05 <andythenorth> partial loading is about station rating
16:58:26 <Eddi|zuHause> _dp_: it does load that. but again, it won't leave if it fills up more quickly
17:00:21 <Eddi|zuHause> _dp_: you can sort of do it with "if load is not 100%, go back to loading station" orders, but that will make the train needlessly move, because the "just stay at the platform" feature is not implemented
17:01:02 <andythenorth> anything about conditional orders needs the 'that escalated quickly' gif :P
17:01:40 <Eddi|zuHause> it's a giant pile of incomplete features with a terribly undesigned UI glued on top
17:01:46 <_dp_> add conditional orders to know bugs :p
17:02:01 <_dp_> just like that: known bugs: 1) conditional orders xD
17:02:18 <Eddi|zuHause> we should start a NoOrders branch?
17:05:15 <_dp_> liberate trains! no one can order them! trains have feelings too!
17:07:23 <nielsm> LordAro: I think my dos music PR should be good now :)
17:20:08 *** Wacko1976 has joined #openttd
17:33:47 <andythenorth> trains go where the cargo is
17:33:50 <andythenorth> no orders needed
17:34:08 <andythenorth> calculate the graph
17:34:18 <andythenorth> calculate the valid refits
17:34:25 <andythenorth> calculate the cargo weighting
17:34:39 <andythenorth> then dice roll for which station, with weighting towards those with more cargo
17:34:53 <andythenorth> then dice roll for a valid destination
17:35:15 <andythenorth> could just be a new conditional order 'any accepting station'
17:35:32 <Eddi|zuHause> i would never want to play that game...
17:35:37 <andythenorth> imagine the chaos
17:50:39 *** johnwhitlow[m] has joined #openttd
18:03:47 *** vorot93 has joined #openttd
18:08:42 <_dp_> hm, that makes me worder, how does cargodist even work with conditional orders?
18:12:15 <andythenorth> fonso had to do things like this 33034
18:31:44 *** Thedarkb1 has joined #openttd
18:38:31 *** virtualrandomnumber has quit IRC
18:39:07 *** snail_UES_ has joined #openttd
18:46:57 *** Wacko1976_ has joined #openttd
19:11:11 *** gelignite has joined #openttd
19:13:11 <peter1138> andythenorth, "also addresses"?
19:13:49 <peter1138> Although, point, having #6145 in the PR title doesn't link it.
19:19:48 <andythenorth> such wasted words
19:19:54 <andythenorth> good job they're not in short supply
19:20:02 <andythenorth> I could debug this nml template
19:20:08 <andythenorth> but it's very sunny isn't it
19:20:16 <Eddi|zuHause> (as in "peak oil")
19:21:20 <Eddi|zuHause> "from this point on, the rate of words produced by humanity will only drop"
19:24:19 <andythenorth> that might be to the good
19:24:35 <andythenorth> I'm not sure we're gaining much from everyone being able to spew words
19:29:52 <peter1138> Did someone decide how to slice water into regions?
19:30:52 <andythenorth> I was trying to figure out a flood fill idea, to handle the irregular edges
19:31:42 <Eddi|zuHause> make it able to slice in any generic way, then let it find out the best way?
19:33:07 <Xaroth> Didn't you have a patch for that, peter1138?
19:34:00 <peter1138> We lost Wolf01 again.
19:35:36 <andythenorth> come back Wolf01
19:38:32 *** HerzogDeXtEr has joined #openttd
19:44:45 <Eddi|zuHause> why don't i have Zak McKracken in my ScummVM?
19:47:11 * peter1138 ponders making his ship locks patch Yet Another Configuration Setting
19:51:07 *** Wacko1976 has joined #openttd
20:05:01 <andythenorth> what does it do? o_O
20:05:58 <Eddi|zuHause> but we can't ever add more settings!!
20:06:05 <Eddi|zuHause> there are too many settings!!!
20:06:21 <Eddi|zuHause> it's even more sacred than the map array!!!!
20:07:24 <andythenorth> there are too many settings
20:07:55 <Eddi|zuHause> there are, but citing that as a reason to not add new ones is crazy
20:11:31 <andythenorth> as long as there's a habit of removing dead settings, then adding new ones is to the good
20:11:43 <andythenorth> it's only a problem if no-one is culling the useless ones
20:12:56 <Eddi|zuHause> remove the 90° setting, make it always on for trains and always off for ships?
20:13:16 <peter1138> Nah, there are people who play with it off.
20:13:33 <Eddi|zuHause> make them see the light! :p
20:13:45 <Eddi|zuHause> my way is the only true way to play the game!! :p
20:14:46 <andythenorth> remove it for ships
20:17:34 <andythenorth> so who approves peter1138's PR pile? :P
20:21:49 <Eddi|zuHause> typical steam: price: 10€, bundle with just that one thing in it: 9€
20:21:55 <LordAro> andythenorth: peter1138 ?
20:22:17 <andythenorth> peter1139 can probably review them
20:22:35 <andythenorth> first rule of fight club
20:22:47 <Eddi|zuHause> don't trust that guy
20:23:09 <peter1138> Eh? No, you can't review your own PRs.
20:25:16 <LordAro> could merge a few though
20:25:44 <LordAro> all are review required
20:27:10 <andythenorth> shall I review them? o_O
20:27:31 <LordAro> some of them have "reviews" from me, but i don't like to tick the approve button
20:27:31 <andythenorth> I think TB would revoke my privileges :P
20:27:36 <LordAro> i don't feel like i have that power
20:27:50 <andythenorth> well there are only about 3 people who can
20:27:59 <andythenorth> it's a bit...slow :P
20:28:17 <LordAro> need to add more people
20:28:22 <andythenorth> it's peter, frosch or TB
20:28:47 <andythenorth> probably RB too, if he has github login
20:28:55 <andythenorth> and alberth same
20:37:51 <LordAro> at the very least we could make more people reviewers
20:39:40 *** tokai|noir has joined #openttd
20:39:40 *** ChanServ sets mode: +v tokai|noir
20:42:53 <Wolf01> Uh nice, zellepins for TF
20:55:52 *** Thedarkb-X40 has joined #openttd
20:58:35 <andythenorth> making me a reviewer :o
20:58:41 <andythenorth> I wouldn't recommend that :P
21:21:36 <LordAro> peter1138: admittedly i didn't even try compiling and just assumed it would work
21:22:36 <peter1138> Or { } depending on how you feels.
21:29:51 *** Wacko1976 has joined #openttd
21:31:34 <Wolf01> andythenorth: should I push the explained TODOs to your repo?
21:31:50 <andythenorth> nah, that one's dead
21:31:54 <andythenorth> peter1138: already made fixes
21:32:03 <peter1138> I've got them somewhere.
21:32:07 <andythenorth> either we fork peter's, or he applies them ;)
21:33:17 <LordAro> or he gives you commit permissions
21:55:21 *** Thedarkb-X40 has joined #openttd
21:56:37 <frosch123> LordAro: what is a runtime-length string?
21:59:50 <frosch123> ah, only SLE_VAR_STRB used length
22:03:57 <LordAro> frosch123: pointer rather than array
22:08:05 *** Wormnest has joined #openttd
22:12:30 <frosch123> last time i commented that the code changes behaviour
22:12:52 <frosch123> now the commit is more complicated to read and still changes behavious
22:16:17 <Wolf01> "Disagree here - the condition is unchanged, it's just moved to an early return style instead of a boolean, which ultimately reduces the length of the code." <- LordAro that seem something Samu would say :(
22:19:04 <frosch123> why is that PR so broken?
22:19:15 <frosch123> everytime i add a comment it hides it as outdated
22:55:34 *** Supercheese has joined #openttd
23:05:16 <peter1138> frosch123, dunno, doesn't do that for me.
23:06:23 <frosch123> yeah, our OnTick stuff does not fit the usual threaded timers :p
23:27:59 *** Supercheese has joined #openttd
23:42:44 *** chomwitt has joined #openttd
23:53:51 *** synchris_ has joined #openttd
23:56:45 <peter1138> Just had the game crash on startup with the ICU issue, like, half a dozen times in a row o_O
continue to next day ⏵
