IRC logs for #openttd on OFTC at 2015-01-24
⏴ go to previous day
00:01:43 *** FLHerne has joined #openttd
00:10:27 <Samu> there is a bug with the listing of vehicle number post 99 in the depot
00:10:55 <Samu> it shrinks from 6x4 size to 5x4
00:15:23 <Samu> I suck at explaining things. Example: build road vehicle 100 -> the grid-list of vehicles changes from 6x4 to 5x4 and the width shrinks
00:21:10 *** FLHerne_ has joined #openttd
00:29:33 *** luaduck is now known as luaduck_zzz
00:35:17 *** Flygon_ has joined #openttd
00:37:35 *** Flygon__ has joined #openttd
00:52:44 *** tokai|noir has joined #openttd
00:52:44 *** ChanServ sets mode: +v tokai|noir
00:53:40 *** shirish has joined #openttd
03:06:28 *** Biolunar_ has joined #openttd
05:02:53 *** Alberth has joined #openttd
05:02:53 *** ChanServ sets mode: +o Alberth
05:06:03 *** tycoondemon has joined #openttd
05:11:15 *** Supercheese has joined #openttd
05:56:16 *** Eddi|zuHause has joined #openttd
06:19:36 *** Flygon_ has joined #openttd
07:11:41 <supermop> hows it going Alberth ?
07:12:51 <Alberth> awake too early, fighting git, snow outside, otherwise spiffy :)
07:14:38 <supermop> was going to go surfing but all of the rental cars in the city are sold out so just played openttd, bought tortillas pork and fruit at the market, got some beer and more openttd
07:15:31 <Alberth> hmm, -1 celcius here at 8 am, going up to 5 or so I believe
07:15:56 <Alberth> lots of openttd can't be bad :)
07:17:08 <supermop> small map with busy bee, full firs (no room for all industries at map creation so i've had to fund a few), pipe and sbb set
07:17:11 *** HerzogDeXtEr1 has joined #openttd
07:17:50 <supermop> and the beer is a new one with some pretty good graphic design
07:22:50 <supermop> good, although it does feel a bit pointless? less pointless that a regular game though
07:23:13 <supermop> on this small map its hard to resist connecting industries other than the goals
07:23:32 <Alberth> :) it's still sandbox play with bb
07:23:55 <supermop> if i get a goal to deliver bauxite, i feel i should go ahead and deliver the aluminum to get more money
07:24:12 <supermop> although i am resisting the urge
07:24:34 <supermop> but after 10 years there are few possible goals not already served
07:24:46 <Alberth> I found it gives out the same goals after some time
07:25:02 <supermop> maybe its better on a huge map
07:25:14 <Alberth> probably the script is too picky
07:25:37 <Alberth> or it should improve on analyzing what's already there
07:25:42 <supermop> i like the idea of ending up with a big legacy network after some years that results from all of these small routes you had to build
07:26:02 <Alberth> on a small map, the lower bound on the distance may be too large
07:26:02 <supermop> rather than planning the comprehensive network from the start
07:26:32 <Alberth> andythenorth and I had the same experience :)
07:26:48 <supermop> as it is challanging to make the lines you built for that first clay goal now make sense for 4 other services as well
07:27:05 <supermop> do you terminate service after the goal is met? i don't
07:27:15 <supermop> maybe i should to receive new goals
07:27:44 <Alberth> local distance = GSBase.RandRange(200) + 50; // Distance 50 .. 250 tiles. <-- that's its aim
07:27:55 <supermop> i like using the sbb sets expensive an slower trains as well
07:29:00 <supermop> with this as i am in the 1930s and i carefully cascade steam engines to new branches when a goal opens up, and replace the mainline service with a $$$ electric
07:29:36 <Alberth> I also keep the service running afterwards, it's more work to take it down again
07:29:38 <supermop> and forcing myself to timetable around 66kmh crocodiles is more fun than just having everything at 100kmh
07:30:00 <supermop> if money were tighter i would move the trains elsewhere
07:31:24 <supermop> the map only has three towns and ive only had one passenger goal though
07:31:47 <Alberth> I like playing with few industries which makes the number of useful lines quite small perhaps
07:31:50 <supermop> and the mainland with most industries has only one town
07:32:02 <supermop> so i am missing passenger transit
07:32:12 <supermop> and the towns seem to languish a bit
07:32:17 <Alberth> pax is one cargo, compared to 30(?) non-pax cargoes :)
07:32:46 <Alberth> perhaps it should allow tuning pax importance in some way?
07:33:00 <supermop> true but i aim to improve the quality of life for all those poor souls growing up in a town with no roads to anywhere!
07:33:37 <supermop> well i usually focus mostly on pax because i feel industries are not urban-focused enough in most games
07:33:46 <supermop> this game i am focused on industries
07:34:06 <Alberth> bb does filter cargos without production, but that fails with firs, as it makes all industries produce something
07:34:42 <supermop> but i thought maybe there could be a town related reward for goal completion though
07:34:50 <supermop> extra money is worthless
07:35:10 <Alberth> that's why you don't get that :)
07:35:12 <supermop> but maybe a short boost to town growth if a pax goal is met
07:35:52 <supermop> although i like the idea of being paid only for my goals
07:36:07 <supermop> thats not sandbox really though
07:36:21 <Alberth> yeah, but it means changing openttd itself
07:36:36 <Alberth> as currently you cannot control that
07:36:51 <Alberth> the script doesn't even know how much money you get
07:36:57 <supermop> well there is the cashdrain method of just clawing back payments
07:37:00 <Alberth> it only knows the amount
07:37:43 <Alberth> yeah, but it's an ugly work around, imho
07:37:45 <supermop> i guess you would need a gs+newgrf
07:38:09 <supermop> newgrf set payment rate to zero or very small, then gs gives you money?
07:38:44 <Alberth> basecost mod to reduce payments
07:39:07 <Alberth> sounds like a nice experiment :)
07:39:49 <Alberth> oh, there was a topic about cargo payments in the forum I think
07:40:16 <supermop> the gs would even work without
07:40:57 <supermop> without it just gives you huge rewards of extra money, with you only get money from gs
07:41:04 <Alberth> you'd be back to giving you money for the goal, which we just established as not-useful
07:41:49 <supermop> and it could pay in various ways though rather than just distance*tile
07:42:33 <Alberth> I still have the idea of having contracts to deliver built into openttd
07:42:48 <Alberth> too bad cargo-dist breaks such things :(
07:42:53 <supermop> me too but i assume no one wants to put that in
07:43:16 <supermop> can gs ask you if you want a goal?
07:43:41 <Alberth> it can open a query window, I think
07:43:54 <supermop> what if it asks if i want to deliver coal to x
07:44:15 <Alberth> there is also the story book, I don't know what you can do there
07:44:47 *** sla_ro|master has joined #openttd
07:45:28 <Alberth> I do want to add such a thing, but as usual I have too many things that I am doing and I want to do
07:45:32 <supermop> i say yes, then it grants me 100 million
07:46:11 <supermop> then i try to run a service for less than that
07:46:23 <supermop> i keep the left over or take a loss
07:46:39 <supermop> and if i fail the goal gs takes the 100M back
07:46:48 <Alberth> that would be the idea, indeed
07:47:15 <Alberth> but you need full control over payments to push a user into that
07:47:30 <Alberth> currently you can just add a random line to make money
07:48:13 <Alberth> thus gs needs to be able to change payment policies
07:49:00 <supermop> this map would really benefit from articulated ships
07:49:22 <Alberth> like I said, I have lots of things I want to work on :p
07:49:35 <Alberth> it would make ships much more fun
07:50:00 <Alberth> I have been playing with this idea with trains, but it's not really working nicely
07:50:20 <Alberth> you cannot dynamically change the train orders
07:50:38 <supermop> auto refit to more than one cargo would also be good
07:51:20 <Alberth> it does that, doesn't it (each wagon can change to a different cargo afaik)
07:52:18 <Alberth> I should experiment more with it perhaps
07:52:21 <supermop> if i want it to carry engineering supplies and building materials one way, but lumber and farm supplies on the way back
07:52:37 <supermop> i can only say refit to lumber or refit to farm supplies
07:53:01 <Alberth> yes, you need more flexible orders
07:53:24 <supermop> if i do refit to any available it could work, but only if CD has already established that both of those cargoes go that way
07:53:45 <Alberth> the current system pushes all options into one order line, which is bound to becomes extremely complicated
07:54:19 <Alberth> that needs to change into a more modular system
07:54:39 <Alberth> having several lines at one station, in some way
07:54:49 <Alberth> *order lines, that is
07:55:01 <supermop> heres a thing that bugs me:
07:55:25 <supermop> often i upgrade before an engines is too old
07:55:44 <supermop> because i need a bigger one on that route, or a bigger ship or whatever
07:56:00 <supermop> and i like to reuse the old one elsewhere
07:56:27 <supermop> but autoreplace just sells the old vehicle, even if it has life left in it,
07:56:40 <supermop> leaving me to buy a new small one for the small line
07:57:10 <supermop> I'd like a 'cascade' option as part of autoreplace
07:57:36 <supermop> where it replaces the old vehicle with a new one, but does not sell the old one
07:57:47 <supermop> instead leaves it in the depot with no orders
07:58:47 <supermop> so train 1 has a steam locomotive but after 3 years i need it to be faster, so it goes to depot and switches for a new crocodile
07:59:03 <Alberth> I understand the idea :)
07:59:09 <supermop> leave with same train number, same timetable, new locomotive
07:59:26 <supermop> but the old steam locomotive is now train 2 and has no orders
07:59:33 <Alberth> autoreplace should learn about consists, imho
08:00:15 <supermop> as it stands i have to do it by hand, which is ok except for lining up with the timetable in a shared group
08:01:31 <Alberth> adding an option in the autoreplace window not to sell the old engine shouldn't be too hard
08:01:54 <Alberth> that is, until you have several replacements going on, and you only want it for some of the old engines, etc
08:03:41 <supermop> well i can always sell the extra old ones
08:03:42 <Alberth> one step further, you want to use the old engine for auto-replacing of another consist
08:04:00 <Alberth> and then it gets REALLY complicated :p
08:04:32 <supermop> autoreplace with "SH 25" - use old one if one is in depot, otherwise buy a new one....
08:04:50 <Alberth> or "wait until one arrives"
08:05:11 <Alberth> number of options explodes somewhat :)
08:05:40 <supermop> or autoreplace and tell the old one to drive to a depot where someone is waiting for it.... hehe
08:05:41 <Alberth> but you'd have the concept of "idle rolling stock" in some way
08:06:18 <supermop> i dont mind just having a pile of old engines and figuring out what to do with them myself
08:08:41 <supermop> i like this game as my little branch lines and dockrailways have an odd assortment of mixmatched old stock from various places
08:09:28 <Alberth> yeah, autoreplace isn't really designed to handle that, currently
08:10:09 <Alberth> I don't even bother with groups, and just do a global auto replace of engines
08:14:44 <supermop> what have you been working on with trains?
08:18:04 <Alberth> in the game, or in programming?
08:18:59 <Alberth> in programming, nothing really
08:19:34 <Alberth> mostly doing GUI things, like adding hide buttons for vehicle purchases, presets that you can save now
08:20:10 <Alberth> currently working on the "new game" window, but a bit stuck on the MHL changes that are coming
08:21:01 <Alberth> in the game, I normally focus on industries, trying to make a "move everything to everywhere" concept for all industrial cargoes
08:21:13 <Alberth> I never succeed though :)
08:21:31 <Alberth> I can't play it for very long
08:21:35 <supermop> container trains everywhere?
08:22:24 <Alberth> that would be the simple solution of course
08:22:58 <Alberth> but also somewhat boring, except for the insane amounts of cargoes you'd have to move
08:23:46 <Alberth> I am also trying to understand newgrfs, but nfo language is so terribly unreadable
08:24:36 <Alberth> the nml compiler is written in terms of nfo concepts as well, which is not helping if you don't understand nfo :p
08:25:49 <Alberth> but I do things like writing Eints, writing busy bee, coding in FreeRCT, and dabbling in the Corsix-TH code fighting with GIT
08:26:20 <Alberth> and a lot of chatting / reading forum :)
08:29:46 <supermop> is it possible to shade certain sprites based on slope?
08:30:17 <Alberth> each slope is a different sprite, afaik
08:30:36 <supermop> so trees on a NW W or SW slope get darkened by x%?
08:30:56 <Alberth> ah, you mean other sprites than the slope :)
08:31:06 <supermop> as it is now, trees completely camouflage the terrain
08:31:09 <Alberth> no idea at all, sorry
08:31:30 <supermop> and there is no way to see if an area is hilly or flat if covered in trees
08:31:46 <Alberth> you need a newgrf guru for that question
08:32:00 <Alberth> seems like a nice idea if it can be done
08:33:43 <supermop> maybe i will post a suggestion
08:36:05 <Alberth> it mostly depends on whether you can make a nice preview of the world quickly
08:37:44 <supermop> well it could just look like the minimap right?
08:39:15 <Alberth> 1px per tile wouldn't work for 4096x4096
08:39:51 <Alberth> but just like the minimap, mostly, is the idea
08:40:10 <supermop> yeah, minimap can zoom out though
08:40:43 <supermop> i guess you'd want tan px for desert and white px for snowy areas
08:41:21 *** KWKdesign has joined #openttd
08:41:26 <Alberth> idea is to first make the bare landscape, do a preview, until happy
08:41:46 <Alberth> then add all other things, like towns, industries, canals, etc
08:41:56 <Alberth> the latter part takes a long time
08:42:50 <Alberth> I was thinking just heightmap, but desert/snow could be useful too
08:43:37 <Alberth> first problem is however to put the entire map in a very small area, yet be able to see how it looks
08:44:26 <Alberth> pixel colours are not difficult to change :)
08:46:33 <Alberth> because of this, I was toying with the original map generator, and found it offers very different game play than terra genesis
08:47:01 <Alberth> lots of small raised areas with little gaps every where
08:54:32 <supermop> i like to play heightmaaps but spend so long correcting them that i lose interest before i even start to play
08:58:25 <Alberth> play an "old" heightmap instead of making a new one :)
09:01:02 *** Flygon_ has joined #openttd
09:01:32 *** andythenorth has joined #openttd
09:13:43 * andythenorth is out of industry IDs
09:14:10 *** TheDude has joined #openttd
09:14:19 <Alberth> full firs is too full :)
09:15:26 <andythenorth> this is because multiple economies
09:15:34 <andythenorth> I can manually re-used IDs carefully
09:15:41 <andythenorth> but eh, it needs ficing fixing properly
09:18:36 <andythenorth> the list of FIRS fixes grows :)
09:18:51 <andythenorth> incomplete templating
09:19:59 <andythenorth> I can sit and poke at it after this release
09:19:59 <andythenorth> first….fix the makefile
09:19:59 <supermop> Alberth: but then i am stuck with new grfs
09:19:59 <andythenorth> requiring ‘make clean’ every time is not helpful :|
09:19:59 <Alberth> supermop: with a heightmap???
09:20:34 <supermop> need to do finescale work with valleys and rivers in SE
09:21:20 <supermop> andythenorth: can i take the sand/clay bunkers from firs industries and throw them on chips tiles?
09:21:37 <supermop> would look better than random piles everywhere in the station
09:21:51 <andythenorth> supermop: probably
09:22:02 <andythenorth> I’m not adding them to CHIPS right now
09:22:11 <andythenorth> got enough things up in the air
09:22:21 <andythenorth> but yeah, if you want to make your own grf
09:22:53 <andythenorth> or fork CHIPS and provide a patch on a ticket
09:23:12 <supermop> pipeline from lime kiln to alu plant looks funny
09:23:38 <supermop> waiting for a coal goal to get this thing going though
09:24:07 * andythenorth going to do a pipe grf in future
09:24:26 <andythenorth> that sounds like super-realisms, a pipeline for lime slurry
09:24:30 <supermop> needs conveyors and log flumes
09:24:51 <supermop> or to people riding logs
09:25:14 <supermop> log has 0hp and cant go uphill
09:26:09 <andythenorth> major problem with fake pipe transport (using trains) is that you have to build double track for the empty return journey
09:26:15 <andythenorth> which is frankly just boring :P
09:27:43 <supermop> im doing ok with just arrow shuttling back and forth
09:28:03 <andythenorth> works for short routes
09:28:13 <supermop> i love how this forum is serving me ads for snack food conveyor systems
09:28:19 <andythenorth> whatever I design, I’m going to try and support single track pipe routes
09:28:43 <andythenorth> probably has to use ridiculously high speed
09:28:51 <supermop> should the pipe even be that viable over long routs
09:28:51 <andythenorth> high capacity is undesirable, takes too long to load
09:29:30 <andythenorth> oh, nmlc hates me
09:29:34 <supermop> i mostly use when the oil well is so close to the refinery it just looks silly to have a train
09:29:48 <andythenorth> FIRS is actually going to hit a hard limit on industry tiles, because nmlc won’t permit double-assignment on them
09:30:03 <supermop> previously i used heqs trams in tunnels hidden behind the refinery
09:30:08 <andythenorth> or maybe it’s just the string identifiers, not the numerics
09:30:16 * andythenorth thinking out loud sorry
09:30:30 * andythenorth -> obligations, bbl
09:35:19 <supermop> i guess its more accurate to supply the aluminum plant with chemicals derived from stone rather than oil
09:37:07 *** JacobD88 has joined #openttd
09:40:38 <Alberth> I don't think accuracy is one of the design goals of firs :)
10:04:51 *** gelignite has joined #openttd
10:30:27 *** Pensacola has joined #openttd
10:47:15 *** andythenorth has joined #openttd
10:47:41 *** itsatacoshop247 has quit IRC
10:57:30 *** tokai|mdlx has joined #openttd
11:11:27 *** Quatroking has joined #openttd
11:13:55 <andythenorth> is red industry minimap colour good for water industry?
11:13:59 <andythenorth> short on good options
11:18:31 <Alberth> green with red is more problematic for many colour-blind people
11:20:44 * andythenorth tries dark green
11:20:51 <andythenorth> might be rubbish
11:25:38 <Alberth> full colour industry colours!
11:37:26 <andythenorth> animated industry colours!
11:48:44 *** HerzogDeXtEr has joined #openttd
11:49:12 *** zeknurn` has joined #openttd
11:53:28 *** zeknurn` is now known as zeknurn
11:54:44 <andythenorth> I should really fix this makefile
11:55:45 <andythenorth> so should make rm generated/pnml/firs.pnml when called?
11:56:08 <andythenorth> oh, the issue is that make sees no changes when called any time after first run
11:56:38 <andythenorth> so if it only works when it finds changes in firs.pnml, I should remove firs.pnml on each make?
11:58:56 <andythenorth> currently I have to call ‘make clean && make install’ which is really inefficient, because it deletes all caches
11:59:14 <andythenorth> so compiles are about 20-40s longer than they need to be
12:04:16 <Alberth> where does firs.pnml come from?
12:04:56 <Alberth> scripts/preprocess.py generates it?
12:05:36 <Alberth> you're not editing that python script ?
12:06:51 <andythenorth> very very rarely
12:07:03 <andythenorth> it used to be a phony target iirc, which worked fine
12:07:14 <andythenorth> but apparently this method is the correct one so I changed it
12:07:42 <Alberth> generated/pnml/firs.pnml: scripts/preprocess.py <-- that means that "generated/pnml/firs.pnml" needs to be rebuild when "scripts/preprocess.py" is changed
12:08:04 <Alberth> so what are you changing instead, and what needs to be done then?
12:08:06 <andythenorth> so it’s behaving as expected
12:08:08 <andythenorth> but not usefully
12:08:35 <andythenorth> I change the things in src/industries, src/cargos, src/graphics
12:08:39 <Alberth> well, it assumes the list dependencies behind the : is complete
12:08:43 <andythenorth> and more rarely, global_constants, and the compile scripts
12:09:26 <Alberth> time to update firs checkout :)
12:10:01 * andythenorth reads history of own changes
12:10:04 <andythenorth> maybe I was being stupid
12:10:52 <Alberth> maybe add ".nmlcache" in the .hgignore file?
12:11:53 *** Myhorta has joined #openttd
12:11:59 <andythenorth> would that cause the makefile to not clean it?
12:12:11 <Alberth> don't know what the script does, so cannot judge that change
12:12:31 <Alberth> no, it prevents hg from listing all cached files as untracked :)
12:12:40 <andythenorth> doesn’t do that for me
12:12:44 <andythenorth> ah, different nml versions?
12:13:02 <Alberth> it is about hg, not about nml
12:13:25 <Alberth> maybe you have ~/.hgignore with a .nmlcache entry
12:13:31 <andythenorth> I have no .nmlcache :)
12:13:44 <Alberth> hmm, that also helps :p
12:13:45 <andythenorth> I have firs.grf.cache
12:14:04 <andythenorth> I am on ancient nml, 0.3.0
12:14:19 <Alberth> frosch changed cache stuff quite dramatically iirc
12:15:14 <andythenorth> I also need to update my nml
12:15:33 <andythenorth> which means a hybrid python2/3 compile, but I have that working in Squid
12:16:15 <Alberth> what in src in actual source, and what is generated?
12:16:39 <andythenorth> 99% of src is src, or 100%
12:16:42 <andythenorth> I’m just checking
12:16:58 <andythenorth> nothing generated in src
12:17:08 <andythenorth> I got rid of that silly mistake :)
12:17:48 <Alberth> but .py and .pnml are both source?
12:18:56 <andythenorth> preventing the .pyc would probably be…sub-optimal :)
12:19:10 *** KenjiE20 has joined #openttd
12:20:52 <andythenorth> are you thinking of doing something like src/* as deps list?
12:22:10 <Samu> hi! issue: i was leveling a huge block of terrain in the scenario editor from a height of 255 to 0 and the game went zzzz. Waited 30 minutes and it was still not over and gave up, I forced close openttd.exe
12:23:49 <Samu> leveling up was not as intensive as leveling down
12:25:10 <Alberth> INDUSTRY_PY_SRC = $(wildcard src/industries/[a-z]*.py) collects all industry .py files
12:25:29 <andythenorth> hmm, I can never curl patches from pastebin and have them apply successfully
12:25:30 <Alberth> INDUSTRY_PYPNML_SRC = $(INDUSTRY_PY_SRC:.py=.pypnml) same collection, but .py -> .pypnml
12:26:07 <Alberth> line 10 and 13 then add both collections as dependencies to the target
12:26:49 <Alberth> if you change any of the src/industry/*.py[pnml] files, the target will be rebuild
12:31:05 <andythenorth> takes a while :) slow compile
12:32:07 *** Supercheese is now known as Guest3076
12:32:12 *** Supercheese has joined #openttd
12:37:47 <andythenorth> Alberth: appears to work, I haven’t tested if it’s picking up changes to deps properly
12:37:53 <andythenorth> before patch: 3m12s
12:37:58 <andythenorth> after patch: 2m 35s
12:38:08 <andythenorth> because the caches are intact and primed
12:38:46 <andythenorth> serious speed up
12:39:32 <Alberth> the other files can be added in a similar way
12:40:11 <andythenorth> plausibly I could remove preprocess .py and call the python render scripts directly from make
12:40:31 <andythenorth> I’m reticent because I fail to make progress with make every time I try
12:40:59 <andythenorth> and also because it’s fundamentally a monolithic compile, no partial compiling, so checking deps is of very low value
12:41:10 *** Suicyder has joined #openttd
12:46:40 * andythenorth wonders how the rest of it works
12:47:04 <andythenorth> the graphics aren’t deps for the .nml file, so if only graphics have changed, no need to process all the nml deps
12:49:06 <Alberth> you need to check being up-to-date of the deps wrt the nml file
12:49:19 <Alberth> or rather, the other way around
12:49:33 <andythenorth> so first I need to make sure all the other .py files are caught
12:49:38 <andythenorth> including src/cargos
12:49:54 <andythenorth> and then figure out how to handle the graphics deps
12:50:16 <andythenorth> urgh there are random legacy files still needed
12:50:22 <andythenorth> like basetiles.pnml
13:05:23 <Alberth> no .pypnml files there?
13:05:35 <andythenorth> properly templated
13:05:47 <andythenorth> this would be easier if I un-messed the FIRS structure
13:05:57 <andythenorth> but testing that is too slow to bear :)
13:06:15 <andythenorth> I’ll do the rest
13:06:35 <andythenorth> so I shouldn’t worry about weird special cases in these deps?
13:06:46 <andythenorth> one-off files with no proper home etc
13:07:50 <Alberth> just add all files where the target depends on
13:08:12 <Alberth> the $(wildcard ...) thing is just a short hand for writing all filenames in full
13:08:24 <andythenorth> so I could add a target for docs too?
13:08:59 <andythenorth> baby steps first I think
13:16:14 <andythenorth> probably some of these files aren’t actually deps for the preprocess.py
13:16:27 <andythenorth> but I’m terrified of having a really fragile, detailed deps list
13:19:53 <Alberth> or are you making a new preprocess.py target?
13:20:21 <andythenorth> more likely I misunderstand make
13:21:02 <Alberth> make thinks in terms of files that change due to changes in other files
13:21:08 <andythenorth> there are some pnml files that are untouched by preprocess.py
13:21:21 <andythenorth> yet they will appear in the deps list that causes preprocess.py to be called
13:21:27 <andythenorth> or at least, that’s how I understand it
13:21:35 <andythenorth> untouched / unrelated /s
13:22:28 <andythenorth> I am not worried about that right now, more a question about my understanding
13:22:32 <Alberth> the only thing that's wrong in your reasoning is that preprocess is not a target, it doesn't change
13:23:06 <andythenorth> the target is “generated/pnml/firs.pnml”?
13:23:23 <andythenorth> and scripts/preprocess.py is one of the rule steps for generating it?
13:24:35 <Alberth> generated/pnml/firs.pnml gets rebuild if the scripts/preprocess.py generator script changes, or when one of the .py or .pypnml files changes that you add at the end
13:25:07 <Alberth> the procedure to rebuild the generated/pnml/firs.pnml is then to call the generator script
13:25:14 <andythenorth> ok so my blanket deps rules are including files that can change without needing generated/pnml/firs.pnml to be rebuilt
13:25:24 <andythenorth> but I am electing to not lose sleep over that
13:25:53 <Alberth> with all the .py and .pypnml files are internal in the generator script, you don't need to list them as arguments to the generator script
13:26:32 <Alberth> that causes unneeded rebuilds at worst
13:26:48 <Alberth> worry about it when it happens often :)
13:32:15 <andythenorth> as arguments? o_O
13:32:50 <Alberth> just like c compilers, gcc -o result.exe *.c :p
13:33:24 <andythenorth> ok, I think that’s fine
13:33:26 *** Eddi|zuHause is now known as Eddi|zuHause2
13:33:27 *** Eddi|zuHause2 is now known as Eddi|zuHause3
13:33:30 <Alberth> but importing .py files that you get as filename is not so simple
13:33:32 *** Eddi|zuHause3 is now known as Eddi|zuHause
13:34:05 <Eddi|zuHause> that sounds like a terrible idea
13:34:05 <Alberth> I never looked into that, as I don't trust user-provided input enough
13:34:27 <Alberth> but some people do it
13:34:34 <Eddi|zuHause> but i think it should be easy
13:35:14 <andythenorth> I wouldn’t do it that way if I was doing it :)
13:35:19 <Samu> hi again.I can't find the setting to have cargo waiting in stations without the need of a vehicle
13:35:24 <andythenorth> I’d look up the name in the existing list I have of modules
13:35:29 *** Absolutis has joined #openttd
13:36:08 <Absolutis> I'm trying to move the openttd data folder from my c: drive to another drive
13:36:18 <Absolutis> Could anyone help me with that?
13:36:40 * andythenorth looks up how to give make just a list of known filenames, no wildcard
13:36:55 <Alberth> your windows system does not know how to move a file?
13:37:11 <Eddi|zuHause> Samu: it's called like "deliver cargo only if there is demand"
13:37:13 *** Myhorta has joined #openttd
13:37:27 <Alberth> andythenorth: BLA = a.x b.x c.x etc just a space as separator
13:37:41 *** frosch123 has joined #openttd
13:37:43 <Absolutis> I know how to physically move it, but how to make openttd look for it in that 2nd drive?
13:38:27 <Samu> that setting yes, I can't find it
13:38:28 <planetmaker> Absolutis, the folders where OpenTTD looks for files are described in readme section 4
13:38:42 <planetmaker> there's no config setting. but close to cfg
13:38:52 *** muzafar has joined #openttd
13:40:33 <Samu> i filter with string "deliver" no results, I'm listing expert category and all setting types
13:41:36 <Alberth> deliver is wrong, you never deliver to a station
13:41:53 <Alberth> try station or cargo or so
13:43:50 <Eddi|zuHause> Absolutis: you probably want to set "working directory" in your desktop/menu link that you use to start openttd with
13:45:28 <Eddi|zuHause> Absolutis: alternatively, you could tell windows to move your entire "documents" folder to the other drive
13:45:47 <Samu> tried station, cargo, demand, deliver, and it doesn't come up. It shows other results that aren't what I'm looking for
13:46:23 *** shirish_ has joined #openttd
13:47:31 <Eddi|zuHause> In file included from /mnt/disk2/spiele/OpenTTD/trunk_clean/src/crashlog.cpp:193:0:
13:47:32 <Eddi|zuHause> /usr/include/libpng12/png.h:2659:31: warning: invalid suffix on literal; C++11 requires a space between literal and identifier [-Wliteral-suffix]
13:47:34 <Eddi|zuHause> fprintf(PNG_DEBUG_FILE,"%s"m PNG_STRING_NEWLINE,(num_tabs==1 ? "\t" : \
13:48:07 <Eddi|zuHause> not sure how and where that comes from
13:51:12 <frosch123> Eddi|zuHause: that warning makes little sense, it's within a macro, so there is actually no suffix
13:51:36 <frosch123> anyway, it's libpng, so no worries for us
13:52:21 <Samu> t.t - I liked that feature for doing vehicle vs vehicle comparisons
13:53:05 <Samu> it was nive to have cargo already waiting
13:53:19 <Alberth> Samu: you can change it in openttd.cfg or in the console with a variable that ends with order.selectgoods
13:53:48 <Alberth> it was also nice to get negative rating due to cargo you didn't pick up
14:16:59 * andythenorth ponders sorting out the makefile more
14:18:02 <andythenorth> giving docs own target would seem wise
14:18:16 <andythenorth> but main makefile already has one, with assumptions that don’t match FIRS
14:18:38 <andythenorth> maybe I can over-ride that without breaking it?
14:21:01 <Alberth> you can use a target as dependency elsewhere
14:23:13 *** deniz1a has joined #openttd
14:24:36 <deniz1a> is there a rail signal that is one way but is ignored in the reverse direction? so a do not enter sign?
14:25:31 <deniz1a> or a go to command but in a negative way: avoid this waypoint for example
14:26:36 <Samu> I found a minor bug. Vehicle age isn't counted correctly. I bought an aircraft on 01-01-2051, and it only became 25 years old on 20-01-2076. 19 days difference.
14:26:40 <deniz1a> i would use one way block signal to stop trains from entering in the reverse direction but that also creates another signal block
14:27:04 <andythenorth> Samu: how many leap years in that period?
14:27:17 <andythenorth> probably not that
14:27:28 <andythenorth> that is a shocking difference though
14:27:38 <dreck> deniz1a a normal one-way signal can't be used from the back side so I'm not too sure I understand what you're asking
14:27:42 <deniz1a> maybe the vehicle was traveling really fast
14:29:00 <deniz1a> dreck: normal one way signal works to block trains entering from the reverse direction but in the forward direction it also creates another signal block or divides the existing block into two smaller parts
14:29:57 <dreck> deniz1a are you not using one-way signal in the first place for your one block?
14:30:05 <dreck> you don't need some "two smaller parts"
14:30:48 <deniz1a> i use it and it works for blocking the trains. but i had already set up the signals and signal blocks. it changes the existing signal blocks
14:31:04 <dreck> mind posting a screenshot?
14:31:04 <deniz1a> so each pair of signals create a signal block, right?
14:31:33 <Alberth> a signal at an end of a track is also a block
14:31:34 <deniz1a> it's not about the shape of the tracks actually. wait i'll explain
14:31:40 *** smoke_fumus has joined #openttd
14:32:17 <deniz1a> Alberth: but you don't need a signal at the end of a track, right? it already ends the signal block
14:33:29 <Alberth> deniz1a: that was just a counter-example for your claim
14:33:56 <Alberth> you can easily make other counter examples, by adding junctions
14:34:52 <Alberth> I don't know if your formulation is wrong, or if you have the wrong idea about signal blocks
14:35:23 <Alberth> andythenorth: need glue?
14:35:47 <andythenorth> I have multiple nmlc installs, for valid reasons
14:36:02 <andythenorth> and ‘make install’ provides none of them correctly :P
14:36:38 <dreck> alberth thats why I asked for a screenshot..would make things easier no?
14:36:55 <Alberth> dreck: quite likely :)
14:38:00 <Alberth> andythenorth: hard-code the NML path in the Makefile ?
14:39:12 <Alberth> although your "compiling_firs.txt" seems to mention adding the right nml to the path :p
14:40:10 <andythenorth> currently it’s just broken in my nml checkout
14:40:15 <andythenorth> haven’t got as far as FIRS yet
14:41:35 <Alberth> the former doesn't use the PATH variable
14:42:17 <Alberth> PATH=.:$PATH make if you want to live very dangerously
14:42:34 <andythenorth> I ran make install
14:42:54 <Alberth> (ie add "." as first directory to search for a command)
14:44:15 <Alberth> how does ./nmlc end up in your firs directory?
14:44:24 <andythenorth> this is in my nml checkout :)
14:44:33 <andythenorth> I broke Squid and Road Hog and such
14:44:38 <andythenorth> I should never upgrade nml
14:44:57 <Samu> according to my understanding, leap years are 2052 2056 2060 2064 2068 2072 2076
14:44:59 <Alberth> cd firs-checkout; PATH=nml-checkout make
14:45:33 <Samu> but the place became 25 years old before February 2076, so 6 leap years
14:45:53 <andythenorth> eh, wrong window :)
14:45:56 <Alberth> if you want different nmls for different parts, you cannot use one PATH variable
14:46:17 <Alberth> somewhere you have to state which one you want for each part
14:47:11 <andythenorth> I have all that figured out before
14:47:19 <andythenorth> but with a broken nmlc it’s of no use :D
14:47:41 <Alberth> Samu: so you also go back to the shop when food lasts a few days longer than it says on the package?
14:47:44 <andythenorth> I think I have to go find my backup disk
14:48:33 <Alberth> hmm, theoretically, you could 'up' the nml checkout while building the newgrf :p
14:49:29 <Alberth> (15:26:35) Samu: I found a minor bug. Vehicle age isn't counted correctly. I bought an aircraft on 01-01-2051, and it only became 25 years old on 20-01-2076. 19 days difference. <-- why is it a bug that the lifespan is 19 days longer than advertised?
14:52:37 <andythenorth> spinning disks are slow
14:53:42 <andythenorth> I replaced my python (virtualenv-ed) bin dir with the one from my backup
14:53:54 <andythenorth> so make install must have broken my python somehow
14:53:58 <Alberth> backup test succeeded :p
14:54:36 <Alberth> disable writing in such directories :)
14:56:18 <andythenorth> ha ha new nml warns about animated pixels \o/
15:02:13 <andythenorth> don’t have old stats for Squid, but new nml seems faster
15:04:10 <Alberth> fr0sch has been busy :p
15:39:53 <Samu> ... and the most profitable aircraft is... dependable on the time it takes in an airport, it actually hurts income on some aircraft types
15:43:36 <deniz1a> can you in reality carry cargo using maglev trains?
15:44:18 <deniz1a> doesn't it have to be relatively light?
15:44:45 <Eddi|zuHause> you can, but it's never going to be economically viable
15:44:49 <deniz1a> and are the railway cargo wagons the same as maglev ones?
15:45:39 <Eddi|zuHause> you need closed/airflow optimized wagons
15:46:08 <Eddi|zuHause> also no, rail wagons cannot go on maglev and vice versa
15:47:13 <deniz1a> wouldn't it be better if train depos were like stations but with different graphics? maintanence stations...
15:47:31 <Eddi|zuHause> maybe, but nobody implemented it
15:47:46 <deniz1a> it just looks funny how multiple very long trains fit into a single tile depot
15:50:50 <deniz1a> then you could choose the length of the depos and they would behave like stations
15:51:50 <deniz1a> but then again, these are not actual vehicles, right? they're just representations of where they are.
15:52:09 <deniz1a> because each tile represents ~600 km
15:53:52 <deniz1a> but if that is true then what about 7 tile long trains?
15:54:03 <Alberth> there is no single distance that a tile represents
15:54:26 <Alberth> scale changes depending on what you look at
15:54:32 <deniz1a> game mechanics wiki says this:
15:54:33 <deniz1a> A tile is, for vehicle speed purposes 664.(216) km-ish, 668 km or 415 miles long.
15:55:07 <Alberth> so, scale for vehicle speed is that
15:55:17 <Alberth> scale for vehicle size is different
15:55:26 <Alberth> scale for house is different
15:55:29 <deniz1a> so there are different scales?
15:55:34 <Alberth> scale for track is different, etc etct
15:55:40 <deniz1a> oh yeah i hadnt thought about that
15:55:47 <deniz1a> those would be some giant houses
15:56:15 <deniz1a> so there are scale inconsistancies
15:56:30 <Alberth> that's why 512x512 is enough for SP
15:57:17 <Alberth> game doesn't gain anything if you make long stretches of track
15:57:32 <Eddi|zuHause> "there are scale inconsistencies" is an understatement if i ever heard one :p
15:57:58 <Alberth> there is no game play in making endless stretches of straight tracks
15:58:35 <Alberth> there are no scale inconsistencies, as there is no attempt to use a universal scale for eveything
15:58:47 <andythenorth> Eddi|zuHause: actually I think there might be a freight maglev test line being built in California
15:58:52 <Alberth> it doesn't work for the game
15:58:55 <andythenorth> flat platforms moving ISO containers at low speed
15:58:58 <deniz1a> oh and building long train tracks is really hard
15:59:27 <deniz1a> you have to correct many terrain issues
15:59:54 <Eddi|zuHause> the true breakthrough of maglev will be when there is a room-temperature superconductor
16:00:25 <Alberth> as long as it's not required to work at that temperature, we have one :p
16:00:27 <deniz1a> it's like flying right above the ground!
16:00:56 <Eddi|zuHause> Alberth: well, obviously that's the requirement :p
16:00:58 <deniz1a> but what if there's some power outage?
16:01:26 <Eddi|zuHause> deniz1a: that's the beauty of it, it doesn't need power
16:01:48 <Eddi|zuHause> deniz1a: the power induced by the gravity of the object is enough to lift it up
16:02:10 <Eddi|zuHause> you need a permanent magnet in the levitation platform
16:02:29 <andythenorth> doesn’t that somehow violate a physical principle?
16:02:38 <andythenorth> or does it assume 100% efficiency?
16:03:17 <Eddi|zuHause> well, sort of, because it assumes the permanent magnet keeps its magnetization
16:03:22 <Eddi|zuHause> it's not a perpetuum mobile
16:03:40 <deniz1a> if they don't need power to levitate then that's great
16:04:04 <deniz1a> or at least not much power
16:04:08 <Eddi|zuHause> but with current technology, you need liquid-nitrogen-ish temperatures
16:04:15 <Eddi|zuHause> which is not cost effective by any means
16:04:16 <andythenorth> which requires an input
16:05:16 <Samu> I had an idea! Rather than comparing aircraft vs aircraft, I'm gonna go compare the different airport types and evaluate their efficiency
16:05:43 <Eddi|zuHause> i'm sure someone did that already, but please go ahead
16:05:46 <Samu> which airport wield better profit :)
16:09:09 <deniz1a> what's the point of constructing a headquarters building? does it bring anything?
16:09:24 <andythenorth> Samu: your boredom threshold is actually remarkably high imho :)
16:09:54 <Alberth> a Tycoon needs a place to live!
16:11:37 <deniz1a> i always build my hq as the first thing
16:11:37 <andythenorth> it’s just a thing
16:11:47 <andythenorth> you need to put a flag in the ground
16:11:59 <andythenorth> the game benefits from certain things that aren’t purposeful
16:11:59 <deniz1a> yes, no flag no country
16:12:08 <andythenorth> in fact, we might want more things that are not purposeful
16:12:31 <andythenorth> I wonder if the trend to pure economic or train-based rationalism is helpful to the fun
16:12:44 <andythenorth> it’s not the only trend, but it is a trend
16:13:19 <Eddi|zuHause> the HQ also produces passengers
16:14:00 <Eddi|zuHause> andythenorth: i think overzealous nonrealism is a bad trend
16:14:08 <andythenorth> equally, perhaps
16:14:20 <Samu> i tried to host an impossible scenario once, but then the players still managed to survive by placing HQs :(
16:14:20 <Eddi|zuHause> andythenorth: it may be a helpful corrective
16:14:36 <andythenorth> there is no single trend, given the number of core devs, the absence of a dictator for life, and the way non-core addons are developed
16:14:42 <andythenorth> but there are definitely trends :)
16:16:28 <Samu> btw there is an issue with infrastructure maintenance costs with autocliff
16:16:50 <deniz1a> on one side, game graphics are being improved to make them more realistic but on the other side gameplay is generally not realistic
16:16:52 <Samu> an opponent can raise the cost considerably for the other opponent
16:17:46 <deniz1a> gameplay realism is more important than graphics i think. which is also nice but it's not everything
16:18:09 <andythenorth> Eddi|zuHause: rejecting things solely because they’re realistic is not particularly useful, but yes, a useful antidote to some approaches
16:18:43 <deniz1a> why would something be rejected because it's realistic?
16:19:14 <Eddi|zuHause> Samu: as said previously, there is no technological way to remove all "exploits"
16:19:33 <andythenorth> deniz1a: because reasons
16:19:58 <Eddi|zuHause> deniz1a: realistic gameplay and fun gameplay are sometimes conflicting goals
16:20:02 <Alberth> realistic is not a design goal
16:20:03 <andythenorth> or rather, my bad understanding of how to use the mercurial
16:20:15 <Alberth> don't eat or drink it :p
16:20:17 <andythenorth> I have uncommited local changes, that are not ready to commit
16:20:21 <andythenorth> but I want to push a fix
16:20:25 <Eddi|zuHause> deniz1a: overly realistic games may turn out to be incredibly boring
16:20:27 <andythenorth> but I can’t because I have to pull and merge
16:20:34 <andythenorth> but I can’t merge because uncommitted local changes
16:20:41 <deniz1a> yes that's true sometimes but it wouldnt be rejected just because it's realistic.. there would be other reasons
16:20:42 <andythenorth> but my local changes break the grf currently
16:20:47 <andythenorth> why is it so hard?
16:21:07 <andythenorth> which is bizarre
16:21:08 <Eddi|zuHause> andythenorth: hg diff > tmp.diff; hg revert, hg merge
16:21:17 <andythenorth> that is stone age :P
16:21:50 <Eddi|zuHause> andythenorth: alternatively, make an mq
16:22:26 <deniz1a> openttd also uses git, right?
16:22:42 <Alberth> no, openttd uses trains
16:22:49 <Eddi|zuHause> deniz1a: there is a git mirror of the openttd svn repo
16:23:23 <deniz1a> why dont they migrate to git or mercurial?
16:23:35 <Eddi|zuHause> most people here use hg to develop
16:23:47 <Eddi|zuHause> but svn has some benefits for a central repository
16:23:49 <deniz1a> why do they keep svn?
16:24:17 * andythenorth reverts and bins local changes
16:24:39 <Alberth> no hg diff > patchfile ?
16:25:27 <Alberth> time to make a little light, and some dinner
16:25:50 <andythenorth> I need the extension pm told me about, provides equivalent of git stash
16:25:55 * andythenorth needs to get off hg
16:27:00 <Alberth> no git <-> hg bridge ?
16:27:35 <Alberth> although hg stores more information than git, so that may be a little difficult
16:28:46 <Eddi|zuHause> deniz1a: one of the reasons for not migrating to hg or git is that people won't be able to agree on one or the other
16:29:19 <Eddi|zuHause> deniz1a: the other one is that an svn as "master" repository has a clean incremental revision
16:29:34 <deniz1a> but you can have that in git too?
16:29:55 <deniz1a> is there something you can do in svn that you cant in git?
16:31:10 <andythenorth> it’s not about svn is better
16:31:14 <andythenorth> it’s about hg vs git
16:31:58 <michi_cc> Alberth: There are several converts/importers, but completely bidirectional bridges not so much AFAIK. hg branches/tags work differently than in git while a git merge can have more than two parents and has distinct author and commiter.
16:32:08 <Eddi|zuHause> deniz1a: it's just what i said, the lack of distributedness stuff is what makes svn better as the "master" repository
16:33:44 <Samu> at first glance, it looks like the commuter is very efficient
16:33:56 <andythenorth> I think if I tried a git-hg bridge I’d probably break the hg repo :)
16:34:02 <Eddi|zuHause> Samu: yes, but it requires small airplanes
16:34:48 * andythenorth wishes he’d measured old nmlc speed
16:36:25 <Alberth> michi_cc: ah so there are more differences thus. I know of only one true bridge, a company like GH, aiming at companies, handles the bridge at storage level at their servers
16:37:13 <Eddi|zuHause> Alberth: so it's more like a new dvcs which offers both hg and git interfaces?
16:37:52 <Alberth> yes, and everybody can use either hg or git it doesn't matter
16:38:01 <Alberth> or at least that's what they claim :)
16:38:26 <Alberth> it's a company owned by the PLY author, but I cannot find the company
16:39:18 <michi_cc> There's git-remote-hg for accessing hg repos with git, but I don't know how complete/data-loss-safe it is.
16:39:24 <Samu> aircraft is in flight? now that' simply not true! it's inside depot!
16:40:54 <Alberth> hg <-> remote git is awful
16:41:09 <Alberth> the git semantics creep in everywhere
16:41:37 <andythenorth> nml tip is marginally faster than 2202 which I had before
16:41:44 <andythenorth> it’s not significant so far
16:41:58 <andythenorth> but it seems much faster, because it has more status output
16:42:03 *** shirish has joined #openttd
16:42:26 <andythenorth> nah, this is on Squid and Road Hog
16:42:29 <andythenorth> which are fast anyway
16:43:01 <Alberth> you can\t even have a cup of tea
16:43:10 <andythenorth> annoyingly, last time I tested, there was no specific piece of nml code that made FIRS slow
16:43:24 <andythenorth> I thought I could maybe drop some features for faster compiling, but no
16:44:29 <Samu> is that supposed to happen?
16:44:31 <andythenorth> I tried partial compiles at the nfo / nml level and failed
16:44:50 <andythenorth> my next plan is to limit the input, so compile only a subset of industries and cargos
16:44:56 <andythenorth> that is very likely to help
16:45:16 <Alberth> I don't know if fr0sch has any plans there
16:45:17 <Eddi|zuHause> don't you already have that?
16:45:32 <andythenorth> I have it for the case of single industry
16:45:49 <andythenorth> but it also requires commenting out nml and CPP macros to make it work
16:45:56 <andythenorth> due to coupling between industries
16:46:54 <andythenorth> it is hugely beneficial when doing stuff like setting up sprite layouts, or testing animation
16:48:00 <andythenorth> I have been slightly inspired by a test runner I have been using
16:48:12 <andythenorth> which can take -t arg with list of tests, or -m with whole module
16:48:39 <Samu> lol, an helicopter just intersected the other in the commuter, one was heading to depot, the other was heading to helipad
16:49:30 <andythenorth> I think ‘proper’ partial compiles with linker are game-over until the string IDs are no longer randomly assigned
16:49:45 <andythenorth> and it might be tmwftlb
16:50:48 <Alberth> since partial compilation means you can compile in parallel
16:52:33 <andythenorth> afaict, string IDs were the only significant blocker
16:52:58 <andythenorth> Iron Horse is using the nml-nfo-grfcodec partial compile route
16:53:21 <Samu> im going to report a bug, i think it is a bug
16:53:25 <Eddi|zuHause> but that's only because you didn't properly follow through with my system of defining the string-IDs
16:54:16 <andythenorth> I don’t think it works meaningfully
16:54:30 <andythenorth> or at least, you didn’t like my solutions, which require bringing all strings in to every compile
16:54:48 <Samu> which category do i pick? NewStation?
16:55:08 <Eddi|zuHause> no, because it's not NewGRF related
16:55:19 <andythenorth> it needs all cargos, and a stub of every industry that uses strings in props or cbs
16:55:43 <andythenorth> and a horrible split on some arbitrary nfo
16:55:55 <andythenorth> which might never go wrong, but seems like The Wrong Thing To Do
16:56:20 <andythenorth> also I got bored of being told I was spending more time than I’d save
16:56:45 <andythenorth> there’s only so much irc mocking any person can take before they abandon a project :P
16:57:30 <Eddi|zuHause> i think it was more about you spending more time than was actually necessary to implement it :p
16:57:47 <andythenorth> that’s because I am a mediocre developer
16:57:51 <Eddi|zuHause> because you fell into a pit of dependency tar
16:58:10 <Eddi|zuHause> that was keeping you from doing anything at all
16:58:29 <andythenorth> I still have a branch sitting there
16:58:33 <andythenorth> it’s not far behind
16:59:43 <andythenorth> feel free to merge it and try :)
16:59:54 <andythenorth> the fragile splitting really bothers me though
17:06:14 *** shirish has joined #openttd
17:11:26 *** shirish has joined #openttd
17:12:53 <Alberth> Samu: can you provide a save game as well?
17:13:45 <Alberth> ie we like a quick way to reproduce the error, so we can concentrate on fixing the problem
17:15:21 <Alberth> it is also better at reporting in the sense that you may have a setting specific for this behaviour that you forget to mention
17:15:41 <Alberth> in a save game it's easy to find it
17:16:18 <Samu> oh, that save is not showing the bug, darn
17:16:37 <Samu> the helicopters must be built for the bug to happen
17:16:38 *** shirish has joined #openttd
17:16:54 <Samu> if they go to hangar after being out, the bug won't happen
17:17:17 <Samu> i will create another savegame, brb
17:22:59 <Samu> ok, i guess I should have foolproffed this bug first
17:25:55 <Samu> k, apparently I need 2 airports
17:26:14 <andythenorth> is there a bug in your bug? o_O
17:27:24 <andythenorth> more fun than buggy tests
17:27:30 <andythenorth> buggy tests are boring
17:27:58 <Alberth> non buggy tests are boring too
17:28:12 <Alberth> so many cases to consider
17:28:20 <andythenorth> I only break them, not write them
17:28:26 <andythenorth> as a consumer, I love tests
17:30:37 <andythenorth> all the cases I’d otherwise forget, or not understand
17:30:40 <Samu> okay basically I need 2 airports. 3rd helicopter must not head to the local commuter, its order must be to a distant airport
17:32:00 <Samu> a skip order is needed on the 3rd helicopter before starting all three, and it must be the 3rd to leave the hangar
17:33:02 <Alberth> I hope you write all that in the ticket :)
17:33:32 <NGC3982> Are you guys affiliated with the CorsixTH project?
17:34:04 <Alberth> I code a little there
17:34:22 <NGC3982> I found it this instant and sufferd a minor deja vu.
17:34:55 <Alberth> ok, never played the original
17:35:11 <Alberth> heck, I barely played CorsixTH either :)
17:35:38 <Alberth> I still have to reach the level of an epidemic :)
17:43:25 <andythenorth> FIRS doesn’t like being compiled with python 2
17:43:38 <andythenorth> AttributeError: 'NoneType' object has no attribute 'days'
17:44:00 <andythenorth> .. /bin/bash: let: x=5502 days, 0:00:00 + 65536 * 0: syntax error in expression (error token is "days, 0:00:00 + 65536 * 0")
17:45:53 <andythenorth> dunno what that does
17:46:24 <NGC3982> Alberth: I'm just about to try the Origin DOSBOX version out.
17:46:25 <andythenorth> not understanding yet, but it’s a python call in makefile
17:47:07 <Alberth> NGC3982: that's not CorsixTH :p but close enough :)
17:47:32 <Samu> heh, the screenshot was actually "telling me" that the 3rd helicopter was going to another airport, dumb me.
17:48:12 <Samu> so is the bug happening there?
17:48:28 * andythenorth bandages this with python2
17:48:36 <andythenorth> but not sure what that call is doing
17:48:44 <andythenorth> Squid et al don’t use it
17:48:58 <Alberth> computes the number of days since 1-1-2000
17:49:07 <Alberth> so you have an incrementing number
17:49:23 * andythenorth wonders how the other grfs are doing that
17:49:27 <Alberth> BB uses findversion.sh for that
17:50:03 <Alberth> although I don't know how findversion does it :p
17:50:10 <andythenorth> maybe Squid just doesn’t have that feature
17:51:03 <andythenorth> I kind of wonder about the FIRS makefile, I think it’s been around in various forms since 2009 or so
17:51:09 <andythenorth> through a lot of changes
17:51:54 <andythenorth> with primed caches, FIRS build time is…
17:51:57 <andythenorth> place your bets?
17:52:10 <andythenorth> previous was 2m 35s after fixing makefile, fairly consistently
17:52:44 <andythenorth> oh that’s for User
17:52:51 <andythenorth> Real is a bit slower
17:53:06 <Alberth> it needs to load python files :)
17:53:21 * andythenorth tests with no cache
17:54:13 <andythenorth> frosch123: the recent nml changes cut Squid and Road Hog compile times by maybe 5%
17:54:19 <andythenorth> but FIRS compile time by 50% :o :)
17:54:50 <Alberth> it does graphics more efficient, but vehicles are too small to benefit much, iirc
17:55:15 <andythenorth> with non-primed caches, User is 2m 11s, and Real is 1m 43s
17:55:38 <frosch123> well, there is still potential left :p
17:55:48 <Alberth> this user is living in the future :p
17:56:18 <frosch123> actually i am surprised it speeds up firs
17:56:26 <frosch123> i thought it only affects x4 sprites
17:56:53 <Alberth> make bigger sprites to get more speed :p
17:57:22 <andythenorth> can anyone else replicate my results?
17:57:29 <andythenorth> previous tests were against r2202
17:58:04 <andythenorth> new numbers are from 2267
17:58:21 <andythenorth> I need to push the combined python2/python3 FIRS makefile
17:59:31 <Alberth> oh dear it needs chameleon and friends
17:59:42 <andythenorth> I’ve pushed makefile change
17:59:59 <andythenorth> wonder if any of the difference is purely down to python3?
18:00:36 *** FLHerne has joined #openttd
18:01:16 <Alberth> it's a pyramid thing? I have 3 chameleon packages
18:01:40 <Alberth> python3 is slower than python2, I think
18:01:50 <Alberth> hg hasn't switched due to that
18:05:56 <NGC3982> There is something i hate with GitHub
18:06:06 <NGC3982> There is never an actual "Hey, download this by pressing this button"
18:06:27 <andythenorth> reading some presentations, python 3 is on average faster or the same
18:06:32 <NGC3982> Took me like two years of linux before i realized i wasn't supposed to.
18:06:37 <andythenorth> but some things benchmark significantly slower
18:09:59 <andythenorth> maybe I could move my python compile scripts to python 3 also
18:10:23 <andythenorth> oh that’s interesting
18:10:40 <Alberth> I run parallel make jobs, and one job needs a file produced by another job
18:12:00 <andythenorth> ok so docs need a target
18:12:13 <andythenorth> I left a comment in Makefile.in about why I did it this way
18:12:45 <andythenorth> I think I should remove preprocess.py, it just makes calls to other python scripts
18:12:53 <andythenorth> makefile should do that
18:13:01 <Alberth> it's not something you discover easily yourself
18:13:46 <andythenorth> I need to put the kids in the bath soon
18:13:55 <andythenorth> then look at adding a doc target
18:25:20 <andythenorth> Alberth: does it work with -j1? o_O
18:29:04 <andythenorth> in Makefile.in, does it work if you remove ‘$(DOCS)’ from the deps for firs.pnml?
18:31:29 <andythenorth> it shouldn’t though
18:32:43 <Alberth> indeed, you have too few dependencies, removing more won't fix that :)
18:32:58 <Alberth> unless you are removing parts of the output
18:33:08 <andythenorth> I think it’s only an accident that it’s working for me
18:33:55 <andythenorth> it works, but I don’t understand why
18:34:05 <andythenorth> unless it happens to be accidentally the correct execution order
18:35:31 <Alberth> it's a race condition, no way to say who will win
18:35:46 <Alberth> is this what you wanted for one nml version?
18:38:17 <andythenorth> yay, /me found a reliable way to trip up make on missing docs
18:39:01 *** LadyHawk has joined #openttd
18:41:16 *** Myhorta has joined #openttd
18:43:03 <andythenorth> so for the cases of primed cache, old and tip are about the same for you? o_O
18:46:05 <andythenorth> I have very poor performance compared to yours
18:47:15 <Alberth> model name : AMD A10-6800K APU with Radeon(tm) HD Graphics
18:49:27 <andythenorth> don’t know how to compare that with mine :)
18:49:46 <Alberth> your's is way prettier than mine :)
18:50:24 <andythenorth> I suspect my first results reflect an older python 2 nml that FIRS was using
18:50:28 <andythenorth> too much string here
18:53:33 * andythenorth tries calling python doc generating script directly
18:53:48 <andythenorth> there is a doc target in the makefile but I’m scared of it tbh
18:54:17 <andythenorth> it’s doing sed and such
18:58:33 <Alberth> {{GRF_TITLE}} {{GRF_ID}} {{REPO_REVISION}} {{FILENAME}} <-- seems to replace these strings by the actual values in the documentation
18:59:35 <Alberth> and some unix / windows line endings stuff
19:01:30 <andythenorth> none of that is used in FIRS
19:05:48 <andythenorth> wow I have a lot of realsprites to fix :)
19:05:52 <andythenorth> need to set ANIM flag
19:06:59 <Alberth> you don't want to give nml the pleasure of finding an animated pixel? :)
19:08:57 <andythenorth> I do love those errors
19:09:37 <Alberth> ha, the ships move in the vehicle list of a dock, depending on whether they are moving themselves :)
19:10:06 <andythenorth> could probably fix that :P
19:10:28 <Alberth> the front wave is a few pixels left of the ship, so the ship moves to the right
19:15:03 <Supercheese> Nothing to fix, I like that feature
19:15:21 <andythenorth> ho ho, nml adds useful animated pixels check, and I immediately suppress it
19:42:42 <andythenorth> so I need a docs target that depends on a lot of things
19:43:00 <andythenorth> anything changes, docs should change
19:45:27 <andythenorth> target can’t be a wildcard list, that would be meaningless?
19:54:43 <andythenorth> can a target be a dir?
19:55:17 <Alberth> the problem I told you? that's the other way around, something needs the generated docs before it exists
19:55:52 <andythenorth> it’s the existing docs handling in the makefile
19:55:59 <Alberth> one solution can be to always build it
19:56:00 <andythenorth> which has only wrong ideas about how to build the docs
19:56:16 <andythenorth> making the docs build correctly, independently, is in theory trivial
19:56:25 <andythenorth> but I don’t understand the existing stuff
19:56:50 <Alberth> I can have a look tomorrow
19:56:55 <andythenorth> and I need to use a directory as a target :)
19:57:27 <andythenorth> suggest it can be done
19:58:13 <Alberth> but that's about the directory itself, not its contents
19:59:50 <andythenorth> contents are all known by the python script that creates them
20:00:03 <andythenorth> it could even write out a manifest for make, but that’s maybe circular :)
20:00:12 <andythenorth> ok, I had to modify the Makefile
20:00:16 <andythenorth> I think that’s bad, but eh
20:01:18 <andythenorth> oh seems to work :O
20:01:25 * andythenorth might have understood a little more
20:02:08 <andythenorth> turning out to be a productive day :)
20:08:22 <nbeee> wed like to create a server for connected gaming
20:09:21 <FLHerne> nbeee: Are you trying to run a dedicated (headless) server?
20:09:24 <nbeee> Please tell us how to do it
20:09:30 <FLHerne> Or host from the normal client?
20:09:46 <DorpsGek> FLHerne: OpenTTD uses TCP and UDP port 3979 for server <-> client communication, UDP port 3978 for masterserver (advertise) communication (outbound), and TCP port 3978 for content service, a.k.a. BaNaNaS (outbound)
20:10:16 <FLHerne> In either case, unless you're on the same LAN, you need to configure your router to forward the above ports to the hosting computer
20:12:24 *** oskari89 has joined #openttd
20:13:17 <FLHerne> nbeee: In that case, what exactly goes wrong?
20:13:41 <FLHerne> What steps do you take, what does/doesn't happen?
20:13:48 <nbeee_> somehow i was logged out
20:13:54 <nbeee_> we are on the same lan
20:14:04 <nbeee_> but we cant create a server
20:17:23 <nbeee_> I could create a server, but its offline....
20:17:41 <nbeee_> nobody can connect it, neither myself
20:18:04 <nbeee_> I tried to create an "internet" server, but I couldnt
20:18:28 <FLHerne> nbeee_: That sounds like it might be a firewall issue
20:18:49 <FLHerne> On the computer, can't be the router if you're on the same LAN
20:19:10 <FLHerne> Are you sure you're trying to connect to the right IP?
20:22:21 <nbeee_> I shut down my firewall now
20:22:35 <nbeee_> We are coneccted to the same IP
20:23:01 <nbeee_> We tried with version 1.4.4, now I am in the version 1.1.4
20:23:58 <nbeee_> two options are different: 1.4.4: advertised Yes or no, 1.1.4 I can choose once LAN or Internet
20:25:04 *** Myhorta has joined #openttd
20:25:44 <FLHerne> nbeee_: Ah, both client and server must be the same version
20:27:23 <nbeee_> sure, we do, we both tried with 1.4.4 first, but didnt work
20:28:40 <FLHerne> nbeee_: If connecting on the same LAN, client needs to connect to the server's LAN IP
20:28:55 <FLHerne> Rather than external IP of the router
20:29:00 * andythenorth is now tempted to refactor all of FIRS
20:29:03 <andythenorth> instead of releasing :P
20:29:22 <FLHerne> nbeee_: Or am I misunderstanding "We are coneccted to the same IP"
20:31:41 <nbeee_> "If connecting on the same LAN, client needs to connect to the server's LAN IP" yes, we did it too
20:58:57 <Samu> is commuter, other than helipads
21:28:14 *** gelignite has joined #openttd
21:31:46 *** zeknurn has joined #openttd
21:32:50 <andythenorth> Guano mine as source of chemicals, too weird
21:32:57 <andythenorth> in combination with the destination
21:34:36 *** itsatacoshop247 has joined #openttd
21:41:48 *** TheDude has joined #openttd
21:45:09 *** andythenorth has left #openttd
22:00:10 *** luaduck_zzz is now known as luaduck
22:19:58 *** itsatacoshop247 has quit IRC
22:34:58 *** minimoo_ has joined #openttd
22:35:09 *** deniz1a has joined #openttd
22:35:45 <minimoo_> 1.5 beta, "Error: Assertion failed at line 141 of /bamboo/src/core/math_func.hpp: min <= max", when trying to start a dedi server instance - known issue or i doing something wrong? ;/
22:35:59 <deniz1a> is there a mod that lets you easily build diagonal train tracks by automatically terraforming the ground?
22:37:31 <FLHerne> deniz1a: ctrl-drag works for terraforming tools
22:37:36 <FLHerne> deniz1a: But otherwise, no
22:38:20 <deniz1a> what does ctrl+drag do?
22:38:33 <Eddi|zuHause> minimoo_: we probably need your .cfg
22:38:58 <minimoo_> Eddi|zuHause: well, that could be issue - i used config from 1.44
22:39:09 <Eddi|zuHause> that should not be an issue
22:39:16 <Eddi|zuHause> but we need it, so we know your settings
22:39:50 <deniz1a> ok holding ctrl lets you do it diagonally, thanks
22:41:12 <deniz1a> and there is no way to build diagonal tunnels and bridges, right?
22:44:38 <minimoo_> Eddi|zuHause: I guess i might need to raise a bug report with that
22:45:04 <Eddi|zuHause> minimoo_: that is probably best
22:48:00 <ST2> sidenote: BTPro.nl- #XX - 1.5.0-beta1 TEST was started with a 1.4.4 cfg file and no issues till now - only needed a shutdown to save version in cfg (that's used by server controller software)
22:48:50 <Eddi|zuHause> minimoo_: i can reproduce the crash with your .cfg, but not with mine
22:51:14 *** Jyggalag is now known as Sheogorath
22:51:16 <glx> oh maybe you can get a call stack then ;)
22:51:28 <minimoo_> it's something in the misc section
22:51:30 <Eddi|zuHause> hm. it doesn't reproduce in gdb :p
22:52:52 <Eddi|zuHause> ah, there it goes
22:53:36 *** Martin89 has joined #openttd
22:54:09 <minimoo_> so yea, removing the linke resosultion resovles
22:54:36 <Eddi|zuHause> minimoo_: fine. but it shouldn't crash with that line anyway
22:55:31 <minimoo_> probably trying to set a resolution on a windowsless display or something :P
22:55:52 <Eddi|zuHause> minimoo_: no, it's trying to resize the window to fit in that resolution
22:56:24 <minimoo_> yea, which for a server running in -D there's no window right?
22:56:48 <Eddi|zuHause> well, there's still lots of stubs for window handling, it's just not drawn to a screen
22:56:53 <glx> internally the game has it's own windows
22:57:43 <minimoo_> in any case, I guess I can play and I suspect you guys will probably have fixed in an hour if you are dev's at least :P
22:58:21 <Eddi|zuHause> the devs that are likely to dig into that are probably asleep
22:58:46 <Eddi|zuHause> but we have enough information to reliably reproduce it, so it should go its way...
23:03:18 <Eddi|zuHause> minimoo_: for the future, you can attach the .cfg as file, you don't need to paste it into the report
23:03:54 <minimoo_> both were on a remote server so being lazy, pasting was actually easiest ;p
23:04:17 <minimoo_> but yea, point taken
23:04:52 <Eddi|zuHause> you can attach links to remote files as file
23:08:59 <krinn> I think TOWN_ACTION_BUY_RIGHTS add an issue to gameplay
23:09:46 <krinn> i've just check, buying rights for town X will disallow anyone to even work with industries the town have authority on
23:10:32 <krinn> just like human, the problem is when you comes with an AI implementing it, it's not human that need to click on each town/each year
23:10:47 <krinn> but a robot that could buys rights of every town in day
23:11:41 <krinn> and the gameplay gets ruins then (you simply have no town to work with)
23:11:43 <glx> but another company can still buy exclusive rights
23:12:40 *** zeknurn has joined #openttd
23:12:55 <krinn> i've just tried, when you brought them, the old rights remains, maybe next year is for you
23:13:32 <krinn> but you will have to do it for every towns, every years
23:14:03 <krinn> (in the meanwhile, all work with that town gets nothing)
23:16:43 *** luaduck is now known as luaduck_zzz
23:52:31 *** itsatacoshop247 has joined #openttd
continue to next day ⏵
