IRC logs for #openttd on OFTC at 2019-07-18
⏴ go to previous day
00:03:44 <Samu> v->targetairport is not updated when the aircraft starts executing next order. It is only updated after takeoff
00:12:12 <Samu> aircraft lands, enters terminal, v->targetairport becomes the airport its at
00:13:13 <Samu> i send aircraft to hangar, what do I have now?
00:24:46 <Samu> I have next_dest equal to last_dest only when the aircraft is in the process of goin up from heliport, that small tiny interval
00:25:31 <Samu> if automatic service requests hangar in this phase
00:27:16 <Samu> v->targetairport is not yet updated, but v->last_station_visited already is
00:27:30 <Samu> I think I can trust v->last_station_visited
00:28:58 <Samu> but v->last_station_visited can be null
00:29:56 <Samu> v->targetairport can be the airport where the plane is built
00:32:09 <Samu> what if i dont visit stations, what if I'm stupid and visit hangars only between airports
00:35:28 <Samu> this code can't be reached in that case
00:36:11 <Samu> seems to be only reachable between heliports,no hangars
00:51:42 <Samu> screw it... last_station_visited it is
00:52:53 <Samu> seems that the code I had, was prepared for the eventual case airplanes would search for hangars outside their orders
00:53:27 <Samu> they don't in current openttd
00:53:51 <Samu> only search on the takeoff
01:09:10 <DorpsGek_II> [OpenTTD/OpenTTD] SamuXarick updated pull request #7104: Fix #5405: Aircraft could route to depots outside their range https://git.io/fhKsL
01:13:51 <DorpsGek_II> [OpenTTD/OpenTTD] SamuXarick updated pull request #7104: Fix #5405: Aircraft could route to depots outside their range https://git.io/fhKsL
01:20:08 <Samu> I have a problem with GetNextStoppingStation
01:23:26 <Samu> what I actually want is a GetNextStoppingStationOrDepot, but this doesn't exist
01:24:33 <Samu> GetNextDecisionNode only considers depots where vehicles are refitting
01:48:08 *** y2kboy23_ has joined #openttd
02:20:54 *** Thedarkb-T60 has joined #openttd
02:22:37 *** y2kboy23_ is now known as y2kboy23
03:08:49 *** rocky11384497 has joined #openttd
07:40:11 *** Thedarkb-T60 has joined #openttd
07:59:05 *** andythenorth has joined #openttd
08:34:42 <andythenorth> seems to find FMSP and ENSP fine
08:35:05 <andythenorth> VEHI is in the cargo table, and defined in the savegame
08:43:04 * andythenorth wonders if cargo_type clamps to first n cargos in CTT?
08:44:28 <andythenorth> 'cargo_type_in_veh' : {'var': 0x47, 'start': 0, 'size': 8},
08:49:16 <Eddi|zuHause> you're probably using it wrong
08:50:00 <andythenorth> unfortunately, I can't find the wrong
08:50:36 <andythenorth> it will be obvious after it's found :P
08:50:49 <Eddi|zuHause> that's how these things usually work :p
08:52:10 <andythenorth> L115 and L116 appear to be working, L117 doesn't
08:55:14 <andythenorth> if I add other arbitrary cargos, they work
08:57:38 <Eddi|zuHause> then you're probably using the wrong cargo
08:58:28 <andythenorth> cargo has cargo_label='VEHI'
09:01:33 <andythenorth> last byte of vehicle var 0x47 is 17, which corresponds to VEHI position in CTT
09:02:26 * andythenorth guesses what it could be
09:02:40 <andythenorth> CTT has VEHI twice
09:04:59 <Eddi|zuHause> it should maybe complain about that?
09:05:45 <andythenorth> nml maybe should
09:05:52 <andythenorth> I'm going to at least write a guard locally
09:07:36 <DorpsGek_II> [OpenTTD/nml] andythenorth opened issue #40: No guard against multiple CTT entries for same cargo https://git.io/fj1rr
09:18:19 <andythenorth> sure :) It will be a few minutes :)
09:26:08 <Eddi|zuHause> not sure if "fix" is the right word
09:34:33 <Eddi|zuHause> also, maybe it makes sense to keep the first value?
09:36:15 <DorpsGek_II> [OpenTTD/OpenTTD] nielsmh merged pull request #7657: Fix: Narrowing cast in one storybook command https://git.io/fj1z9
09:37:07 <DorpsGek_II> [OpenTTD/nml] andythenorth commented on pull request #41: Fix #40: issue a warning on duplicate CTT entries https://git.io/fj1oL
09:42:14 <DorpsGek_II> [OpenTTD/nml] Eddi-z updated pull request #41: Fix #40: issue a warning on duplicate CTT entries https://git.io/fj1rA
09:42:33 <Eddi|zuHause> andythenorth: with the new change, your old code should work correctly
09:43:16 <DorpsGek_II> [OpenTTD/nml] Eddi-z commented on pull request #41: Fix #40: issue a warning on duplicate CTT entries https://git.io/fj1oc
09:45:46 *** andythenorth is now known as Guest7857
09:45:46 *** andythenorth has joined #openttd
09:46:03 <Eddi|zuHause> andythenorth: saw the new change?
09:46:26 <Eddi|zuHause> andythenorth: your old code should now work correctly
10:16:13 *** andythenorth has joined #openttd
10:16:36 <andythenorth> Eddi|zuHause: makes sense to do that
10:16:48 <andythenorth> does it need testing? Looks like it will just work
10:17:00 <Eddi|zuHause> andythenorth: would be nice to have confirmation
10:17:17 <Eddi|zuHause> both that it does what it says, and that it didn't break anything (obvious)
10:21:10 <DorpsGek_II> [OpenTTD/nml] andythenorth commented on pull request #41: Fix #40: issue a warning on duplicate CTT entries https://git.io/fj1KI
10:22:20 <peter1138> andythenorth, yes, I have a branch for newgrf-defined docks.
10:22:29 <peter1138> andythenorth, it is incomplete.
10:32:44 <Eddi|zuHause> andythenorth: now we only need to find someone who can merge it :)
10:33:42 <Eddi|zuHause> peter1138: does that need some indication which tiles can be used for docking, and which ones are just decoration?
10:34:17 <Eddi|zuHause> peter1138: that could also solve the "ships dock at the wrong end of the default dock if a river happens to run past" situation
10:34:49 <peter1138> It needs someone to finish it, basically.
10:34:59 <peter1138> I could try getting back into things again at some point.
10:35:14 <peter1138> Maybe taking my yoke & pedals away from the desk would help :p
10:35:34 <peter1138> And removing bikes.
10:35:45 <peter1138> Although apparently it's going to rain this weekend, so maybe.
10:36:23 <peter1138> It suffered the usual problem, somewhat your fault, of "but State Machien!111"
10:38:41 <peter1138> Maybe I should finish that too, along with Eddi|zuHause's patch for it.
10:39:22 <Eddi|zuHause> state machines are always my fault, it seems
10:39:54 <peter1138> Did vehicle variants need a state machine? I can't remember...
10:44:24 <reldred> CS layperson here; what's the aversion to state machines?
10:45:13 <peter1138> Designing them with desired features and flexibility, almost set in stone from the start, to be used via NewGRF byte-code.
10:45:50 <peter1138> Most of the time the feature requirements are "1) unicorns 2) is it merged yet?"
10:46:01 <reldred> Yeah I guess you're trying to define a finite set of states and a finite set of inputs/putput?
10:49:20 <Eddi|zuHause> reldred: the two problems are: A) deciding what features should be accessible, B) how to specify stuff in NFO but also keep the programmer safe, and C) how to handle deadlocks
10:49:38 <reldred> it's like scoping projects with customers; list of deliverables: 1) shit we haven't paid for 2) delivered yesterday
10:50:24 <reldred> Keep the programmer safe as in physically safe from harm from deranged NFO coders and users who all want competing things? :P
10:51:12 <Eddi|zuHause> safe from mental breakdown
10:51:30 <Eddi|zuHause> because it will be impossible to debug
10:51:30 <reldred> So the problem not so much is with state machines themselves but more that you can only really build them once and knowing the end users are never satisfied with what they've got
10:51:59 <reldred> yeah more states and more inputs means I guess an ever growing matrix of possible combinations.
10:52:26 <reldred> and possible ways of offending the userbase for $REASONS
10:52:54 <reldred> but the latter I suppose applies to everything I guess
10:54:37 <Eddi|zuHause> reldred: say you're programming a bus station, with different parallel loading bays, you want a feature to load-balance the incoming busses across the bays
10:54:48 <Eddi|zuHause> so you need a path reservation feature
10:55:06 <Eddi|zuHause> and the ability to read which paths are reserved or occupied
10:55:31 <Eddi|zuHause> and a feature to check whether the (articulated) bus is too long for the bay
10:56:20 <reldred> Yeah righto. THat makes good sense actually because I'm looking at the turnstile state machine example on wikipedia
10:56:38 <reldred> So yeah, suddenly it's starting to spiral out of control a bit
10:57:53 <Eddi|zuHause> reldred: or you're programming a drawbridge over a canal. you suddenly need to specify separate paths for the rail across and the canal underneath, blocking one path or the other, and setting signal states, and stuff
10:59:06 <Eddi|zuHause> and that's still just the A) part
10:59:25 <Eddi|zuHause> and we're already hoplessly drowning in features
11:00:35 <reldred> And I guess because OpenTTD is still very much adhering to the TTDPatch ideology of every feature being disableable to an extent. You've then got the interaction of logic between features that are also seperately enabled and disabled.
11:01:29 <reldred> or more trying to compartmentalise it so it doesn't interact
11:06:02 <andythenorth> peter1138: it's not like there's a deadline, do it when it's interesting :)
11:18:56 <reldred> ahhh, rainfall produces some really pretty maps
11:19:23 <reldred> I need to sit down and try and get the diff to apply against jgr's patchpack
11:19:44 <reldred> rainfall plus everest would make for some really nice maps.
11:23:12 <Eddi|zuHause> whenever i mess with river generation, i come to this barrier that river tiles can't go on all slope tiles
11:25:32 <Eddi|zuHause> you have to work around them, or terraform them, but in either case, it messes with your algorithm
11:26:14 <reldred> Rainfalls algorithm has its shortcomings, it seems to be croaking entirely on my heightmaps at 1024x2048, but it can stomach 512x1024
11:26:49 <reldred> I get the feeling if I looked under the hood more it'd be very... expensive to run. Certainly blows out map generation times severely.
11:26:54 <reldred> And doesn't take much to break it.
11:28:43 <reldred> general question, but would openttd be able to stomach the concept of a backwards foundation? Instead of extending vertically up and draw foundations underneath to instead extend down and draw foundations against the tear created against the landscape?
11:29:29 <reldred> I think for rivers for instance it wouldn't make sense to draw foundation under it
11:30:48 <Eddi|zuHause> reldred: you could go that one step further and allow arbitrary height cliffs. main problem with that is lack of map rotation to see behind mountains
11:31:36 <reldred> Yeah, I remember those sort of convos. At least with only one step of height difference there is still the ability to partially see the obscured tile with transparency options.
11:31:50 <reldred> Perhaps a transparency option to hide the landscape tiles themselves?
11:32:21 <peter1138> reldred, rainfall looked quite nice. If you make the diff apply against master instead, then merging it with JGR becomes JGR's problem ;)
11:32:31 <Eddi|zuHause> reldred: one idea was "slicing" the terrain at some heightlevel
11:32:50 <reldred> Eddi|zuHause: I've seen other vidya games do exactly that
11:32:59 <reldred> Can't remember exact examples though
11:33:23 <reldred> peter1138: that's what I did, but I stuffed something up when I tried to apply it myself.
11:33:26 <Eddi|zuHause> reldred: that's mostly an UI problem, no big technical hurdles
11:34:09 <reldred> do I need to post a bounty? "GiveReldredRealCliffs"
11:34:10 <peter1138> Well, the main thing is to understand it all, fix most of the issues, and get it merged to master.
11:34:22 *** andythenorth has joined #openttd
11:34:39 <reldred> peter1138: Yeah, most of it seems to be pulled into its own files and then referred back from its hooks into the UI, worldgen, etc.
11:34:53 <reldred> But it's all way way way over my head.
11:35:03 <peter1138> Best way to learn... dive in ;)
11:35:41 <reldred> Hah, well my gentle poking and prodding of stevef2 finally got it compiling in VS2019 so I'm happy for the moment ;)
11:38:34 <reldred> Eddi|zuHause: just had a gander through the comments on that one, also as an fyi the guy who picked up rainfall after ic111 does have his branch up on github.
11:39:44 <reldred> Eddi|zuHause: most of the points you raised about yours failing (local terforming to fix dead ends, etc.) are part of rainfall but I also feel like that's where it starts to then get really computationally expensive. it has a whole bunch of steps to widen rivers, fix deadends, prevent uphill flow, find a good destination for the flow, etc.
11:40:42 <reldred> mind you the heightmaps I keep feeding it are cruel and unusual.
11:41:34 <Eddi|zuHause> reldred: i've never actually looked at the rainfall patch, but i imagine the worst case for that is a completely flat area
11:42:29 <reldred> It doesn't do too badly depending on the users settings. Lowlands can sometimes look a bit like the Mississippi but tbh I kinda enjoy sadistic maps so I'm okay with that.
11:43:24 <andythenorth> breadth first anyone? o_O
11:48:17 <Eddi|zuHause> guys, civ4 turns out to be not fun when 4 people gang up on you, and you're 2 tech levels behind
11:50:51 <peter1138> Is it lunch time yet?
11:51:07 <nakki> i grabbed lunch over an hour ago
11:51:11 <andythenorth> DorpsGek: is it lunch yet?
11:51:29 <andythenorth> no answer, now we're screwed
11:56:20 <reldred> you would leave a decision as important as lunch up to a cold uncaring machine? :O
11:56:59 <andythenorth> I substituted elevenses
12:01:06 <Artea> anyone there knows a MMORPG called MU Online ?
12:56:46 *** Trinity^ has joined #openttd
13:57:28 <andythenorth> articulated vehicle carrying wagons?
13:57:38 <andythenorth> somehow articulated wagons never work
13:57:47 <andythenorth> oh it's because they can't be flipped :P
15:33:08 <andythenorth> I thought I could use Road Hog trucks as loads for Horse vehicles wagons
15:33:17 <andythenorth> but they're drawn to same height as trains :P
15:33:24 <andythenorth> so they are way too high
15:34:51 <Samu> I need a GetNextStoppingAirport
15:35:20 <Eddi|zuHause> means you must completely redraw the set? :p
15:38:49 *** Thedarkb-T60 has joined #openttd
15:58:15 <peter1138> Samu, is that something aircraft specific or something missing from orders?
15:58:43 <Samu> because airports are also hangars
16:00:20 <Samu> I'm trying to determine next_dest that is within aircraft maxrange
16:00:46 <Samu> GetNextStoppingStation ignores depots/hangars
16:01:06 <Samu> if the next order is go to hangar, it won't consider it
16:01:31 <Samu> thus it can screw with the max range of the aircraft
16:06:40 <Samu> order 4: Go to the nearest hangar, this order triggers GetNextStoppingStation line I linked
16:06:51 <Samu> next stopping station is A
16:07:17 <Samu> but the order I have below is going to C hangar, it's skipping that
16:23:34 <Eddi|zuHause> "update scheduled: tomorrow 3:56AM"
16:26:08 *** andythenorth has joined #openttd
16:26:15 <andythenorth> Eddi|zuHause: which set? Horse or Hog? Or both?
16:26:22 <Eddi|zuHause> andythenorth: yes
16:26:47 <Eddi|zuHause> andythenorth: although, i meant hog in that case. also, i was being ironic
16:27:21 <andythenorth> I think it's plausible that I could draw trucks that are less tall, as load sprites :P
16:27:45 <andythenorth> I used the Hog sprites as waiting cargo in CHIPS platforms though :D
16:33:08 *** gareppa has joined #openttd
16:41:53 <Eddi|zuHause> you could just fudge a pixel here or there?
16:44:33 <andythenorth> just cut the cabs down
16:52:26 <andythenorth> should vehicle wagons be sloped? :P
17:11:59 <Eddi|zuHause> andythenorth: slopes probably look weird at this scale
17:14:10 <peter1138> Why not draw 1x scale ;)
17:14:26 <peter1138> This old 4x zoom-out stuff... I dunno...
17:25:45 <reldred> they do look good though andythenorth, looks a little weird at that scale but that's just ttd I suppose
17:37:12 *** Beerbelott has joined #openttd
17:41:10 <nielsm> huh, wikipedia and other wikimedia stuff down?
17:47:45 <Eddi|zuHause> a glitch in the matrix?
17:48:14 <peter1138> The Matrix is showing at my cinema tonight.
17:48:25 <peter1138> I can't be arsed to go though.
17:48:50 <Eddi|zuHause> it's probably still a nice movie
17:48:57 <Eddi|zuHause> haven't watched it in like a decade
17:49:08 <peter1138> Probably but I have an SSD to fit.
17:49:43 <Eddi|zuHause> i wanted to screw open my dead SSD, but it has "special" 5-point star screws
17:50:00 <LordAro> Eddi|zuHause: that's because you're not supposed to open it :p
17:50:03 <Eddi|zuHause> i managed to get two of them off, but the third one won't budge
17:50:16 <Eddi|zuHause> LordAro: why should that stop me? :p
17:51:31 <Eddi|zuHause> also, i can never find the exact screwdriver i'm looking for
18:03:04 *** Progman has joined #openttd
18:06:43 <andythenorth> Eddi|zuHause: you don't have a pentalobe screwdriver set? :o
18:07:00 <Eddi|zuHause> i don't have... anything...
18:07:14 <Eddi|zuHause> and even if i did, i wouldn't find it in a situation where i needed it
18:09:18 <andythenorth> anyway, those trucks are too tall, they hit the electric wires
18:18:29 <Samu> should I work on a GetNextStoppingAirport at all?
18:19:36 <Samu> or there's another way to solve the problem?
18:20:48 <Samu> or am I using the wrong function
18:41:11 <nielsm> _dp_: any thougts about the above?
18:45:26 * andythenorth needs to draw better trucks cargo
18:56:08 <andythenorth> or should it be cars?
18:56:21 <andythenorth> hmm the industries and stations show trucks :P
18:59:21 *** Smedles has joined #openttd
19:06:11 *** sla_ro|master has joined #openttd
19:13:27 <Samu> GetNextStoppingStation doesn't do what I think it does grr...
19:15:30 <Samu> it's smart, but not smart enough for what I think it could do
19:16:56 <andythenorth> more random colours needed?
19:17:34 <Wolf01> More random shapes needed
19:19:41 <nielsm> some vans and dump trucks would be a better mix
19:20:04 <nielsm> maybe try the trucks on a car facing each other instead of both in same direction
19:20:30 <andythenorth> I'll try a few things
19:22:00 *** HerzogDeXtEr has joined #openttd
19:27:38 *** sla_ro|master has joined #openttd
19:46:03 *** andythenorth has joined #openttd
20:07:20 *** gelignite has joined #openttd
20:19:58 <Samu> linkgraph doesn't seem to include service at nearest depots
20:20:07 <Samu> or am i seeing things wrong
20:28:58 *** Thedarkb-T60 has joined #openttd
20:34:54 <nielsm> depots are not stations you can pick up or deliver cargo at
20:35:01 <nielsm> hence they don't participate in the link graph
20:40:07 <nielsm> the link graph is about where cargo can get from and to, no where vehicles stop
20:42:23 <nielsm> the link graph is about where cargo can get from and to, no where vehicles stop
20:44:59 <Samu> are you sure? it will stop at refit orders
20:45:41 <nielsm> the link graph is about where cargo can get from and to, not where vehicles stop or how they have the capacity
20:45:54 <nielsm> the link graph REPRESENTS the possible cargo flow
20:46:18 <nielsm> as links between stations the cargo can move directly between
20:46:44 <Samu> *LinkRefresher::PredictNextOrder - if the order is a refit at a depot, it is handling it
20:46:54 <nielsm> I never said it does not consider refits
20:47:09 <nielsm> I said the link graph is not about vehicle order lists
20:47:20 <nielsm> the link graph does not represent vehicle orders
20:47:24 <nielsm> it represents where cargo can move
20:47:54 <nielsm> the mechanism for generating the link graph involves analysing vehicle orders
20:49:06 <Eddi|zuHause> you sound like you're having fun...
20:52:56 <Samu> conditional orders with service at nearest hangar doesn't work too well with GetNextStoppingAirport
20:53:10 <nielsm> conditional orders break everything
20:53:30 <nielsm> you may as well give up as soon as an orders list has conditionals in it
20:54:00 <Samu> if load percentage is equal to 0, for instance, it considers 2 airports
20:54:14 <Samu> but doesn't even evaluate the load percentage :(
20:57:06 <nielsm> you're probably way overthinking this
20:57:46 <Samu> order 3: go to nearest hangar, order 4: jump to order 2 if load % = 0
20:58:11 <nielsm> compare distance to previous and current destination, go to the closer of the two that exists, if closest is out of range break down and go the closest airport regardless of orders list
20:58:20 <Samu> nearest hangar order makes the conditional order not work
20:59:48 <Samu> it kinda does that already
21:00:11 <Samu> current destination is... not perfect
21:00:40 <nielsm> also remember: garbage in garbage out
21:01:05 <nielsm> if the player does something stupid like destroy and airport the aircraft was heading for, just do whatever is easiest for you
21:01:15 <nielsm> don't try to fix the player's mistakes for them
21:01:17 <Samu> if it's a place where it can load/unload ,it is considered
21:01:37 <Samu> if it's a hangar, then... well... it doesn't consider
21:02:11 <Samu> so I'm trying a GetNextStoppingAirport which also considers go to hangar orders as stopping places
21:04:23 <Samu> the stupid thing is 2 orders that are like this: order 1:) go to nearest hangar, then order 2) go to hangar named A, then order 3) go load cargo at B
21:04:38 <nielsm> <Samu> order 3: go to nearest hangar, order 4: jump to order 2 if load % = 0 <-- but where was the aircraft positioned when it reached order 3, it wasn't in mid-air because that's not possible (unless the player presses skip order)
21:04:49 <Samu> GetNextStoppingStation doesn't detect A, only B
21:06:31 <nielsm> you should go read up on what "undefined behaviour" is in the context of C and C++ programming
21:06:50 <nielsm> then consider how it can relate to this kind of problem you're looking at
21:07:24 <nielsm> hint: it has to do with deciding that certain cases are not worth trying to handle or there is no good way to handle them, so you can just do whatever is easiest
21:07:54 <nielsm> player tries to do something dumb => player receives dumb or undesirable behaviour in return
21:08:07 <Samu> in that case GetNextStoppingStation is sufficient
21:10:14 <glx> never try to fix player mistakes, just make sure player mistakes don't cause asserts :)
21:10:54 <nielsm> like, if the player tries to destroy their last airport while they have planes in the air
21:11:17 <nielsm> I don't even know what happens then
21:12:05 <glx> probably invalid orders and the plane goes in "infinite" loop
21:12:23 <glx> and good luck to click on the plane if it's a fast one ;)
21:13:08 <nielsm> it explodes in the air
21:13:35 <glx> it's handled properly I think :)
21:13:49 *** rocky11384497 has joined #openttd
21:13:52 <nielsm> and the broken hull is now hovering 50 meters in the air
21:14:28 <Eddi|zuHause> i remember some screenshot ages ago where someone sent a bunch of helicopters circling, then destroyed the airport
21:15:28 <glx> at least the game engine doesn't break, that's the more important thing
21:15:50 <Eddi|zuHause> i don't even know if that's an original feature
21:16:12 <andythenorth> oof these vehicle transporter wagons are just untidy
21:16:24 <andythenorth> how to make better :P
21:16:33 <Eddi|zuHause> burn them with fire
21:24:06 <nielsm> original ttd behaviour is just to go to the top tile and circle
21:28:31 <andythenorth> the truck sprites are just bad :P
21:28:37 <andythenorth> and the wagon sprites are just bad
21:33:28 <Eddi|zuHause> andythenorth: but that's higher zoomlevel
21:34:18 <Eddi|zuHause> yes, but still too many pixels
21:35:58 <Eddi|zuHause> i recently saw a video on how lego roads changed over time, because vehicles went from a width of 4 to 6
21:36:56 <glx> I think I only saw width of 4
21:37:23 <Eddi|zuHause> it was 4 in the 90s
21:37:48 <Eddi|zuHause> which is basically all the lego i got
21:38:42 <glx> but my old 4.5 motor (usable with trains and wheels/caterpillar) had width of 6 with wheels
21:39:34 <DorpsGek_II> [OpenTTD/OpenTTD] SamuXarick updated pull request #7190: Fix #7188: AI instance crash when reloading AI in a server. https://git.io/fh9jW
21:39:36 <glx> it's from the time trains were 4.5V or 12V
21:39:53 <andythenorth> I took that motor apart
21:40:04 <Eddi|zuHause> glx: i only got 9V monorail stuff
21:40:25 <Samu> i should actually let #7190 die
21:40:25 <glx> I added the 4.5V motor to the push/pull train :)
21:40:52 <Samu> i've only solved a conflict with this push
21:40:53 <Eddi|zuHause> an some lego technic stuff which is way bigger
21:41:20 <glx> some recent technic are huge
21:41:49 <Samu> can u put a stale in there again?
21:42:47 <Eddi|zuHause> andythenorth: the bilevel car wagons might be interesting, but that won't cover trucks/vans
21:45:43 <andythenorth> I guess I just need to do better drawing :P
22:28:03 *** Eddi|zuHause is now known as Eddi|zuHause2
22:41:33 *** Eddi|zuHause2 is now known as Eddi|zuHause
22:45:54 <DorpsGek_II> [OpenTTD/OpenTTD] SamuXarick updated pull request #7193: Fix #6468: Load correct version of AI as specified during the time of its save. https://git.io/fhHI1
22:46:59 <DorpsGek_II> [OpenTTD/nml] michicc merged pull request #41: Fix #40: issue a warning on duplicate CTT entries https://git.io/fj1rA
22:54:54 <Eddi|zuHause> i was thinking whether RTT need a similar treatment, but it's a bit complicated with all the fallback stuff
23:22:59 <michi_cc> LordAro: I can approve it, but I don't know if I have merge rights.
23:23:13 <LordAro> oh, it's glx with admin, isn't it?
23:23:16 <glx> I think I'm the only one with merge rights
23:23:52 <LordAro> and frosch & pm, apparently
23:26:06 <michi_cc> Is there a point to run both 3.8 and 3.9 for the CI? It would increase running time.
23:26:42 <glx> 3.8 is the default on debian
23:28:19 <LordAro> i didn't want to remove 3.8 until i knew 3.9 was working :p
23:28:35 <LordAro> 3.8 is default, but 3.9 is easily available
23:29:05 <LordAro> also technically speaking, default clang on debian stable is now 7.0 :p
23:29:12 <LordAro> (given buster was released last week)
23:29:24 <glx> and by default we still rely on gcc anyway
23:29:43 <glx> 3.8 is just an extra CI check
23:30:11 <LordAro> yeah, and we're now in the situation of gcc9.2 warning about something that clang3.8 requires
23:30:47 <glx> I think adding an image is not really an issue anyway
23:30:58 <LordAro> just adding an image isn't, no
23:31:11 <LordAro> it requires the azure pipelines changing to actually pick it up regardless
23:31:23 <LordAro> (which i've done in #7630 already)
23:32:29 <glx> oh nightly failed to generate the docs image
23:33:05 <glx> error 503, will be ok tomorrow I guess
23:33:37 <LordAro> could it be retriggered?
23:35:21 <DorpsGek_II> [OpenTTD/OpenTTD] SamuXarick updated pull request #7376: More than max_no_competitors could be created in network games https://git.io/fjvY8
23:35:47 <glx> not really important as the bundles were published anyway
23:36:06 <LordAro> yeah, not like they've changed :p
23:38:28 <glx> for now we build 19 images
23:39:06 <glx> most of them being bases for other images
23:40:06 <glx> and release images use gcc only
23:40:33 <glx> so yeah upgrading clang should not be an issue
continue to next day ⏵
