IRC logs for #openttd on OFTC at 2017-09-10
⏴ go to previous day
00:05:07 *** tokai|noir has joined #openttd
00:05:07 *** ChanServ sets mode: +v tokai|noir
00:52:31 *** chomwitt has joined #openttd
01:25:32 *** ToBeFree has joined #openttd
01:26:31 *** chomwitt has joined #openttd
01:52:10 *** Hiddenfunstuff has quit IRC
02:17:46 *** supermop has joined #openttd
02:47:04 *** orudge` has joined #openttd
02:47:04 *** ChanServ sets mode: +o orudge`
03:29:33 *** supermop has joined #openttd
03:49:04 *** orudge` has joined #openttd
03:49:05 *** ChanServ sets mode: +o orudge`
04:32:37 *** orudge` has joined #openttd
04:32:37 *** ChanServ sets mode: +o orudge`
04:45:01 *** Biolunar_ has joined #openttd
05:24:56 *** Montana has joined #openttd
06:27:40 *** tux is now known as Guest4393
07:39:15 *** Alberth has joined #openttd
07:39:15 *** ChanServ sets mode: +o Alberth
07:59:28 *** andythenorth has joined #openttd
08:21:27 <andythenorth> probably time for a BB game eh
09:01:03 *** mindlesstux has joined #openttd
09:10:30 <Alberth> hmm, likely, but busy matching database records against each other in python
09:12:55 <Eddi|zuHause> sounds like you should do that in sql instead?
09:13:34 *** orudge` has joined #openttd
09:13:34 *** ChanServ sets mode: +o orudge`
09:17:09 <andythenorth> something has to generate the sql, no?
09:18:09 <Eddi|zuHause> (i have actually no clue how that works)
09:19:39 <andythenorth> you’re not missing anything
09:24:02 <Alberth> xslt is just fun, you have 3 different languiages (input language, output language, and rewrite-control language), and all of them are in xml
09:25:09 <Alberth> I haven't found how to rewrite several steps in sequence, not sure it is actually supported
09:26:26 <Alberth> Eddi: yes, sql would be nice, if only all records were complete, did not have spelling mistakes, wrong data, etc etc
09:27:34 <Alberth> ie the kind of stuff you can expect when humans copy millions of records from centuries old books
09:27:56 <andythenorth> some new project? o_O
09:28:37 <Alberth> somewhat, just providing some code to construct matches between the records
09:28:51 <Alberth> I am not running the project :)
09:30:05 <Alberth> 9 level deep for loops :p
09:31:51 <Eddi|zuHause> Alberth: the few steps i made towards xslt once made it sound like it was a tree reorganization tool, which made it sound scary enough, and only works on xml, which made it useless enough, to not care any further
09:33:21 <Alberth> I did a few steps more, but once I found you could not concatenate rewrite steps, usefulness reduced to near-zero
09:34:06 <Alberth> ie cat input | xslt step1 | xslt strp2 | xslt step3 > output didn't feel sane :p
09:34:30 <Eddi|zuHause> yeah, that sounds like it would be useful
09:34:53 <Alberth> I still wonder if I just missed it, or if it's really not possible
09:35:07 <Alberth> it seems a very basic notion to have in a functional rewriting language
09:35:49 <andythenorth> in my limited experience of it, xslt is a solution looking for a problem
09:36:07 <Alberth> but as you said, being xml in everything, it gets too complicated too quickly to be of any real use
09:36:52 <Eddi|zuHause> andythenorth: in my experience, that happens a lot in maths. where 100 years later someone comes and finds the problem
09:37:09 <Alberth> andy, it may one of these cases where the problem you're solving is simpler than the solution :p
09:37:52 <Eddi|zuHause> like, this "boolean algebra" thing, that was completely useless
09:37:52 <andythenorth> xslt for rewriting existing HTML -> new DOM
09:38:11 <Eddi|zuHause> ... until someone built a computer
09:38:15 <Alberth> Eddi: that works in math, since you're talking about fundamental properties. xslt is simply avoided by picking any other solution
09:39:18 <andythenorth> xsly is the purpotedly ‘simple’ solution for rewriting html on the fly, because it can be handled by a lightweight stateless WSGI app, simple pipeline
09:39:37 <andythenorth> but the simplicity of the architecture moves all the complexity into the actual transforms
09:39:41 <andythenorth> where it really sucks
09:40:08 <Eddi|zuHause> sounds like you need an xslt generator :p
09:40:21 <Alberth> it heavily depends on how much rewriting you want, I am sure there are valid cases for it
09:40:57 <Alberth> eg simple tag rewriting, or inserting your content somewhere
09:41:02 <Eddi|zuHause> but now, enough mocking of a language i barely understand
09:44:15 <andythenorth> meh, cdist flooded my pax stations :|
09:46:09 <Eddi|zuHause> yeah, it does that :p
09:46:35 <Eddi|zuHause> that's why i never get to do industrial networks
09:47:06 <Alberth> just ignore cities, they're annoying anyway :)
09:47:41 <andythenorth> BB told me to do it :P
09:48:25 <Alberth> :O it created a chain of connections?
09:48:37 <Alberth> it doesn't even aim for that :)
09:52:19 <andythenorth> tol me to deliver pax :)
09:52:53 * andythenorth should stick to point-to-point networks :P
09:57:56 <Eddi|zuHause> point-to-point lines with single trains on each. best way to play the game
09:59:32 <Alberth> depending on your goal of playing :p
10:23:26 *** sim-al2 is now known as Guest4401
10:23:27 *** sim-al2 has joined #openttd
10:45:53 <Eddi|zuHause> now, am i really missing something for this unorganized crime interview, or am i just not getting the rules?
10:48:10 <Eddi|zuHause> i mean, it offers me to skip it, but i solved all the other minigames before...
10:59:16 <milek7_> how to change game script in running game?
10:59:34 *** supermop has joined #openttd
11:00:15 <Alberth> milek7_: not, I think
11:00:31 <Eddi|zuHause> ah, i found it :p
11:00:49 <Alberth> ic111: these start times and offsets don't show up elsewhere?
11:01:09 <milek7_> eh, so patching again
11:01:59 <Alberth> milek7_: there are reasons why you cannot do things
11:02:13 <Alberth> if things are safe to change, we generally allow it
11:11:59 <ic111> Alberth: No, they don't show up elsewhere
11:12:21 <ic111> They are properties of the timetable (aka OrderList), and the vehicle
11:15:13 *** gelignite has joined #openttd
11:16:07 <Alberth> offset isn't really needed, it can be derived from the difference between start times
11:16:24 <Alberth> if you put them under each other, you gain lots of width
11:16:38 <ic111> No, it works the other way round: You specify the offset, and based on this, the vehicle start is calculated.
11:16:55 <ic111> I thought about removing one of (Timetable | Vehicle) start
11:17:23 <ic111> However, if I remove Vehicle start, then players don't have the start date relevant for the arrivals / departures below
11:17:26 <Eddi|zuHause> depending on your use case, you might need one more than the other
11:17:27 <Alberth> only vehicle start would qualify for removal, I think
11:17:57 <Alberth> so there is no vehicle departure anywhere?
11:18:09 <Alberth> ie first one would be equal to start, right?
11:18:39 <ic111> Below, you specify "Departure 3rd April 1905", and do this in terms of the *Vehicle* timetable.
11:19:00 <Alberth> don't think in implementation, user doesn't know or care
11:19:24 <Alberth> oh sorry, you meant something else
11:19:42 <milek7_> ok, game script can be changed in scenario editor
11:20:01 <ic111> Just because I think in terms of users, I allow them to input the departure of a particular vehicle in terms of the timetable of that particular vehicle.
11:20:26 <Wolf01> So, is the departure date vehicle based and tied to the next order? Why don't you just move it on the vehicle status bar (start/stop button) so you always see it at glance?
11:20:27 <ic111> But of course, in the background, I translate the departure into the range of the timetable, i.e. [Timetable start, Timetable start + length [
11:20:31 *** Progman has joined #openttd
11:21:28 <ic111> So, the Vehicle start is a piece of information players need IMHO
11:24:28 <Alberth> sure it's useful, but vehicle start time is not equal to first departure in the table?
11:25:41 <ic111> Alberth: No, in general not.
11:26:09 <Alberth> ok, add a first row that is?
11:26:32 <Alberth> just throwing suggestions
11:27:13 <ic111> I even don't actually enforce that dates are within the timetable range - I once did, but in my experience this just triggers tedious problems once you do things like shifting a timetable, editing start dates, and so on. Instead, I now mark orders with arrivals / departures outside the range [STart, start + length[ red.
11:27:38 <Alberth> I don't think you need "+" and "=". If a user sees the times, and cannot see how it's computed, I think he has no business doing time tables
11:29:16 <ic111> Wolf01: The departure dates exist once per Timetable, but what you see in a vehicle timetable in fact is [Departure in timetable + Vehicle offset]. I don't understand what you mean with "move it to status bar"
11:29:54 <Wolf01> I don't get timetables
11:31:21 <Alberth> time tables make the assumption of a regulated schedule, which may not apply to the play style of a user
11:31:30 <Wolf01> I only use them for the most reliable thing: make vehicles wait n-days at a station
11:31:32 <Alberth> eg I never care about schedules
11:32:04 <Wolf01> Load if available and wait for 5 days
11:33:21 <ic111> A frequent situation for me would be: Load if available, and depart once (given the overall setup of your network / timetables) you know that you have a free slot (i.e. no other train crosses your way) on the railway ahead
11:33:26 <Wolf01> It's the only thing I would save for timetables and move it directly on the orders options
11:36:22 <ic111> That maybe would improve your rating at a station, but it would not allow for setting up a network where you have control about, where your trains meet, where (and when) they quickly exchange freight / passengers, and so on
11:36:51 <Wolf01> I guess you even play with breakdowns disabled
11:37:27 <ic111> Actually, designing such a network with breakdowns on is a quite interesting / challenging task
11:38:00 <ic111> The key idea is that you always specify your departures with some additional time for the potential case of a breakdown in mind
11:38:37 <ic111> plus that you let vehicles wait an appropriate amount of time at the end station, to ensure that they can start on time into the reverse direction
11:39:33 <Wolf01> To me it's a concept that won't work, in my network trains already wait each other, I can have only 1 train per track anyway so they are forced on stations or waiting points/passing loops
11:41:33 <Alberth> breakdowns are not very bad by themselves, they only cause massive havoc if your network is too full
11:41:45 <Alberth> which never happens if you do careful planning
11:42:06 *** Celestar1 has joined #openttd
11:42:08 <ic111> ... and exactly that careful planning is the thing I do using timetables
11:42:30 <Alberth> unlike the usual playing style where you just throw a zillion trains onto one single giant network
11:43:33 <ic111> After all, if you have e.g. five or ten trains sharing the same timetable, how do you (without timetables) ensure that they never by an unprobable, but possible sequence of breakdowns cause a traffic jam at the one end of the network, and the station at the other end isn't serviced for half a year
11:43:44 <Alberth> I don't plan, but observe the train flow, and expand the network when breakdowns start to cause more waiting
11:44:54 <Alberth> expanding the network is a lot of pain with time tables?
11:45:46 <ic111> Do you mean, introducing new railways / stations / trains, or you mean adding additional tracks / platforms?
11:46:45 <ic111> I mean, if I design a railways line as one-track-line, where trains meet at stations, and realize that I need a second track, then I just build it
11:46:46 <Alberth> my main cause is more stations or more trains, which both eventually lead to a busier network
11:47:41 <Wolf01> Also vehicles auto-separation is a weird concept since it's done at the first start, I never managed even to start vehicles on different dates, I can't think about when I need to add more vehicles, I usually do it by hand by starting new vehicles when I find a gap
11:47:45 <Alberth> as I don't plan, the trains will sort themselves out on any additional track
11:48:28 *** HerzogDeXtEr has joined #openttd
11:48:37 *** Wormnest has joined #openttd
11:48:51 <Alberth> but if you plan things, doesn't more trains cause havoc in the tables?
11:48:57 <Wolf01> <ic111> I mean, if I design a railways line as one-track-line, where trains meet at stations, and realize that I need a second track, then I just build it <- that is wrong, if you set the timetables right you won't need a second track even with 100 trains
11:49:39 *** orudge` has joined #openttd
11:49:39 *** ChanServ sets mode: +o orudge`
11:49:46 <Alberth> there is an upper limit to the number of trains you can push onto a track Wolf01 :)
11:49:58 <ic111> More stations / trains: Well, I have fun with planning things, i.e. when I add an additional train with a new timetable, then I look, at which date do I have a gap, where it can depart at the station; or maybe, how do I need to change the departures of the other trains to allow for such a gap.
11:50:00 <Wolf01> Yes, but with timetables you make them wait
11:50:42 <ic111> Wolf01: There is a difference between, what theoretically works, and what works (with breakdowns enabled!) in practice.
11:51:37 <ic111> So, you might plan that your trains meet here and there, and then realize, that some have breakdowns, and that your plan just causes delays in practice (one train has a breakdown, arrives too late at a station, thus the opposite train cannot start on time, and the delays spread over the network)
11:51:55 <ic111> If something like this happens, then extending the infrastructure / more tracks is needed
11:52:00 <Alberth> ah, ok, makes sense ic111, constant planning and replanning :)
11:52:03 <Wolf01> So if you say you can know when a specific train uses a track, you should also know the throughput of your track and timetable trains correctly
11:55:26 <ic111> Wolf01: Breakdowns are a probabilistic concept. Also, stations not always produce the same amount of cargo (i.e. sometimes loading takes a day more or less), so it's somewhat experience-based, and in fact, for me this is part of the fun
11:55:51 <Wolf01> That's what I can't get about timetables
11:56:50 <Alberth> the fun is trying to get as close as you can to the max load on a track, while at the same time keeping all schedules
11:57:28 <Alberth> since the max load isn't a simple number, this is a difficult problem
11:58:47 <Alberth> or rather, challenging
11:59:26 <ic111> Another aspect is, that some delay is often ok, but once some event happens you might get a bigger problem
12:00:23 <Wolf01> For me is: if you use timetables you don't give a fuck about production, you just want your network running like a well oiled mechanism
12:00:38 <Alberth> it's not a simple "run smooth" -> "total chaos" sharp edge, it degrades as you approach the limit
12:00:45 <ic111> E.g., it might be ok, if your long-distance-train has 5 days delay, but once it has 10 days delay, the local train supposed to enter the line after the long distance train instead enters in front of it, and the long distance train suddenly has 30 days delay because it travels behind a slow train
12:01:41 <Alberth> Wolf01: you can care for production, just add more trains and don't use them fully
12:01:59 <Wolf01> Which is what I already do without timetables
12:02:48 <Alberth> nobody claimed that time tables are the only possible solution :)
12:02:56 <Wolf01> At least for vehicles, with trains won't work, with 3+ trains I already fill the network
12:03:48 <ic111> I talk about connected rail networks, where 100 or more local, long distance and freight trains share the same tracks...
12:04:22 <ic111> ... and where a long distance train can only be fast if I ensure that it meets other trains that cause it to wait as seldom as possible
12:05:34 <Wolf01> If only there was a "depart after train 123 arrived at station"
12:07:11 <ic111> Basically, it's an incremental playing approach, I have timetables for all vehicles, the goal is that they are more or less on time, and when I change things, add new railways, add new stations, add new trains, then I (constant replanning) change things in a way that things still work afterwards
12:07:30 <Alberth> Wolf01: more likely you want "a train from blah group"
12:09:14 <Eddi|zuHause> people frequently request "leave if another train of the same group arrived"
12:09:52 <Eddi|zuHause> where "group" might mean "order list"
12:13:26 *** Hiddenfunstuff has joined #openttd
12:14:28 *** orudge` has joined #openttd
12:14:28 *** ChanServ sets mode: +o orudge`
12:20:37 *** andythenorth has joined #openttd
12:23:05 <Wolf01> BTW, I think if timetables will be merged to the orders list and not having their own window, and the single orders could be collapsed/expanded like the station cargo list with cdist enabled, then it might be more usable, yes the UI will be a lot cluttered but you have everything in one place and you can hide the things you won't use
12:23:47 <andythenorth> don’t you like having two entirely separate concepts for controlling vehicle behaviour? :P
12:24:05 * andythenorth still doesn’t understand how waiting for 10 days over-rides the full-load order
12:24:19 <Wolf01> No, because I would like the waiting time at station in the order list and don't use anything else of timetables
12:24:33 <ic111> In fact the timetable window I implemented has easily accessible buttons top-right specifying "Show timetable + Order info", "Show timetable info", and "Show order info"
12:24:49 <Alberth> andy, it needs a toggle somewhere in the orders
12:25:06 <andythenorth> it breaks my brain :)
12:25:11 <andythenorth> I will leave you to it :)
12:25:12 <ic111> But it wasn't meant as complete replacement for the traditional order window in the first place (although all the functionality is available)
12:25:17 * andythenorth playing Busy Bee game
12:25:21 <Alberth> the entire order stuff is beyond limits of what you can express nicely, currently
12:25:27 <andythenorth> please please please don’t make it worse anyone :(
12:26:14 <Alberth> ic111: given the amount of duplicate information, it would make sense to merge them
12:26:31 <Alberth> but the orders need work too, they are too limited
12:26:46 <ic111> As I have said, in terms of functionality merging is possible
12:26:52 * andythenorth still doesn’t understand the purpose
12:26:56 <Alberth> and merging would explode your patch, likely
12:27:03 <ic111> But I understand if people who don't like timetables just want to keep the usual orders window
12:27:24 <ic111> Thus I was very reluctant to do so
12:27:41 <Alberth> it would need some way to hide timing information, basically
12:28:02 <Wolf01> I would like a better orders window, but I can't find a purpose for timetables in my play style
12:28:20 <Alberth> likely there isn't a use for you
12:28:30 <Alberth> but it's a way of playing the game
12:28:41 <Alberth> ie much like RL trains run
12:28:54 <ic111> Basically, the way are the "Full", "Time", "Dest" buttons in the upper right section of my window
12:29:01 <Alberth> I wouldn't use them either
12:29:03 <andythenorth> my view of timetables is somewhat distorted
12:29:16 <andythenorth> I only tried them for ships, and they just don’t work for ships
12:29:28 * andythenorth should open some FS about that
12:29:29 <ic111> Time currently means, "Show only the timetable lines", "Dest" means show only the destination = order lines, and Full means, "show both kinds of lines"
12:29:59 <Alberth> and ships is the simplest, as they can pass eachother
12:34:06 <ic111> Probably, attempting to really merge those windows in this patch is just too much, then I not only have a discussion about timetables, but also one about every single feature of the orders window
12:34:54 <ic111> Though the current situation, that I have a timetable window where all the functionality of the orders window can be reached, can be a good starting point for eventually doing so
12:35:44 * andythenorth wonders if there’s no other way to space vehicles out
12:35:49 *** sla_ro|master has joined #openttd
12:36:12 <Alberth> make an invisible wagon :p
12:36:30 <ic111> It's not only about spacing; but you might read the log of this channel of today, just before you entered it.
12:36:32 <Alberth> A trick I sometimes use is to have a single very long block
12:37:04 <Alberth> or rather just a block equal to the desired distance
12:37:51 <Alberth> extending the orders is going to be very very messy, I think
12:38:17 <ic111> What do you mean with "extending the orders"?
12:38:34 <Wolf01> Mmmh, I think I have a problem... I see the same vehicles twice in the replace vehicles window
12:38:50 <Alberth> multi-cargo transfers at a single station eg
12:39:20 <ic111> You mean specifying which cargo should be unloaded / loaded to another vehicle?
12:39:32 <Wolf01> Oh, I loaded egrvts twice for an unknown reason
12:39:37 <Alberth> only unload coal (but keep the wood), load stones
12:40:12 <Alberth> grf doesn't detect double instantiation :p
12:40:20 <ic111> Such things make the orders line longer and longer, and need GUI work
12:40:46 <Alberth> I think you need to split to simpler orders, and have several
12:41:03 <Alberth> rather than trying to cram everything in a single line
12:41:27 <Alberth> ie loading has nothing to do with unloading or driving to the next station
12:41:27 <ic111> Basically the concept one uses for timetables
12:42:06 <Alberth> so why not have some sequence of small orders that can be done in parallel if it makes sense
12:42:24 <ic111> Basically, we talk about the visual design here
12:42:36 <ic111> not necessarily about orders in a technical sense
12:42:53 <Alberth> I'd do that also in the back-end
12:43:30 <Alberth> or you will end up with some extension, and then again run into limitations as the back-end is not flexible enough
12:44:00 <Alberth> obviously this totally flips all order code upside down :p
12:44:02 <andythenorth> I know we banned UI scripting (I found it an old thread about it in forums recently)
12:44:17 <andythenorth> eh, that might need rethought :P
12:44:27 <andythenorth> timetables would be much better handled out of the core game
12:44:27 <ic111> Things that can be specified multiple times (e.g. your "load that cargo, undload that cargo") are an 1:n relation, that needs redesign in the backend
12:44:33 <Alberth> users now can program, suddenly?
12:45:07 <Alberth> ic111: yep, total replace would it be, I think
12:45:28 <Alberth> takes a few years at least
12:45:59 <ic111> But something like the arrival or the departure is just a value in the Order structure, I don't see why moving this to a different place in the backend should be necessary
12:46:23 <andythenorth> Alberth: they manage to assemble newgrfs :P
12:47:12 <Alberth> sure, arrival and departure are the simple things
12:47:41 <Alberth> although hard-coded override of 10 waiting days... isn't understood by all
12:48:04 <Alberth> so likely there is more room for configuring behavior if you want
12:48:46 <Alberth> andythenorth: yes some do, but that doesn't mean the majority of the users
12:50:28 <ic111> If you really want to approach a total redesign, then you need to agree first on the big picture (i.e., which order properties should be supported in the long term?), and then you can gradually refactor things towards that goal
12:50:47 <andythenorth> the puzzle for me is how to design something with integrity
12:51:09 <andythenorth> where the goals are unclear, and few people understand them
12:51:18 <Alberth> ic111: I know, I re-implemented 120+ windows
12:52:05 <andythenorth> it was fun recently asking players about daylength
12:52:19 <andythenorth> learning was that there isn’t one idea of daylength
12:52:38 <Wolf01> I have 3 ideas of daylength
12:52:46 <andythenorth> so even making daylength technically viable doesn’t meet wishes
12:53:08 <andythenorth> daylength could be done in newgrf, with a handful of factors
12:53:21 <andythenorth> timetables don’t work in newgrf, but eh
12:53:31 * andythenorth wonders if the primary goal can be stated
12:53:49 <ic111> of a potential refactoring?
12:54:00 <andythenorth> just even of timetables
12:54:26 <andythenorth> I know of multiple player goals, but listing those isn’t quite same
12:54:41 <andythenorth> - spacing out buses in town
12:54:46 <andythenorth> - modelling real-life networks
12:54:54 <andythenorth> - optimising network use + flow
12:55:15 <ic111> Speaking of my concept:
12:55:30 <andythenorth> - proxy for ‘partial load'
12:55:56 <andythenorth> - triggering station delivery of cargo on first run of route, to ensure vehicle makes money
12:56:01 <ic111> Spacing out buses can be done - just choose an arbitrary start date, use autofill if you want, or specify departures explicitely, and then build vehicles with different offsets
12:56:57 <ic111> If you have vehicles with offsets 0, 1, 2, 3 months, and the timetable length is four months, and set the departures accordingly (either by autofill, or manually), then you will have buses that run approximately in one month distance
12:57:45 <ic111> Modelling real-life networks: Well, about what I described above; I control which train is where at which time, and where train meet, and when a long distance train enters a line shortly before a local train, and such things
12:58:54 <ic111> Optimizing network use + flow: At least, my attempt is that once a train left at a station it has to stop as seldom as possible at signals
12:59:22 <ic111> But of course, you need some concrete example in order to really talk about this
12:59:53 <ic111> proxy of partial load I don't understand
13:00:03 <Wolf01> The problem with spacing vehicles is that, as I said, you need to fix it continuously by hand, it would be cool that timetable recalculates the times based on the number of vehicles sharing the schedule, but that would conflicts with any other use of timetables
13:00:46 <ic111> Wolf01: Doing this automatically is a very very hard problem in computational sense
13:01:29 <ic111> I mean, I can detect that a train was late, but determining why it was late... hard
13:01:52 <andythenorth> ic111: proxy of partial load <- setting wait ’n’ days is equivalent to load x%
13:02:21 <ic111> Ah, then this basically is no application of timetables
13:02:25 <andythenorth> because load x% isn’t provided, so timetables are ~essential for that (or I have to provide different vehicle choices)
13:02:45 <andythenorth> it’s emergent behaviour, which has become “vital”, as much as anything is vital in a game about pixel trains :P
13:03:08 <ic111> I mean, if you solve this using timetables, then you implicitely assume that loading occurs in a certain speed
13:03:35 <Wolf01> I don't think so, you can estimate the travel time with manhattan distance between orders and average vehicle speed, then divide per number of vehicles, cache it and recalculate only when the number or type of vehicle changes, you can account also for an offset as roads aren't always built perfectly and you might have breakdowns
13:05:12 <ic111> Triggering station deliver of cargo on first run - how is this related with timetables?
13:05:15 <Wolf01> ic111: I'm not speaking of fine tuning the timetable to keep vehicles in sync to the single second and pixel, it's just an estimation of the time it takes, simple math average
13:05:56 <Wolf01> Autofill can go wrong for every kind of reason, I don't even rely on it
13:06:06 <ic111> I also use it quite seldom
13:06:16 <andythenorth> ic111: if the vehicles are set to wait n days on first run, they are less likely to travel empty, so they make money immediately
13:06:31 <ic111> Ok, this kind of optimization
13:07:59 <ic111> Such a temporary change can be done either by temporarily changing the timetable, or by a temporary full-load-phase
13:08:27 <ic111> I personally usually don't care too much about the first round of the vehicle
13:09:34 <Wolf01> I usually set up the orders and forget, I don't want to keep an eye on them because stuff needs to be fine-tuned continuously
13:09:53 <andythenorth> I set buses and trams quite often to wait 5 days or so
13:09:59 <andythenorth> more especially big pax ships
13:10:04 <andythenorth> at every station
13:10:24 <andythenorth> full load is no good, especially on asymmetric routes
13:10:34 <andythenorth> e.g. 500pax ship stopping at a tiny village
13:10:52 <andythenorth> so that is _my_ primary goal for timetables :)
13:12:04 <ic111> If you want that 5 days permanently, you (in my patch), especially for ships which cannot wait at red signals, can use autofill if you want to keep things simple
13:12:33 <ic111> Or, if you don't want to have time permanently you always have to do something once you don't want them any longer
13:18:14 *** ToBeFree has joined #openttd
13:29:23 *** ToBeFree has joined #openttd
13:38:07 <andythenorth> it’s so weird building newgrf airports :)
13:38:11 <andythenorth> non-standard orientations
13:58:32 <Wolf01> <ic111> If you want that 5 days permanently, you (in my patch) [...] can use autofill if you want to keep things simple <- wat!?
14:38:44 *** ic111__ has joined #openttd
14:41:41 *** frosch123 has joined #openttd
15:02:51 <andythenorth> ^ plausible this will be fixed?
15:03:07 <andythenorth> I don’t mind either way, but I need to delete some vehicles from IH if not
15:05:28 <frosch123> i wouldn't know what a correct behaviour would be, unless you want template-based replacement
15:08:00 * andythenorth doesn’t know either
15:08:14 <andythenorth> easiest to avoid it by not creating those vehicles
15:08:21 <andythenorth> codeless code :P
15:09:42 <frosch123> is the report about it being a *slient* failure, i..e do they want a more noticeable alert?
15:10:28 * andythenorth proposes that mixed articulated vehicles are just banned
15:10:45 <andythenorth> they’re not needed
15:25:45 *** orudge` has joined #openttd
15:25:45 *** ChanServ sets mode: +o orudge`
15:34:13 <andythenorth> Iron Horse might be getting un-simple :|
15:44:37 <frosch123> classic would be a wooden horse
15:50:58 <andythenorth> _probably_ I should delete the dual power electro-diesel train
15:51:06 <andythenorth> and definitely not add more of them
15:52:01 <frosch123> why? it's an interesting graphical gimmick
15:53:08 <andythenorth> not sure I’ve ever used it in a game :)
16:02:21 <DorpsGek> Commit by adf88 :: r27913 trunk/src/console_cmds.cpp (2017-09-10 16:02:13 +0200 )
16:02:22 <DorpsGek> -Fix: 'unban' console command was not handling IPv6 adresses properly
16:03:36 <DorpsGek> Commit by adf88 :: r27914 trunk/src/console_cmds.cpp (2017-09-10 16:03:29 +0200 )
16:03:37 <DorpsGek> -Fix/Feature: 'unban' console command - fix invalid help text and be more verbose
16:08:56 <adf88> what do you think about too big national letters (sprite font)
16:09:10 <adf88> do you find any intention in being so big?
16:10:26 *** orudge` has joined #openttd
16:10:26 *** ChanServ sets mode: +o orudge`
16:10:47 <andythenorth> haven’t tried to repro
16:10:53 <andythenorth> dunno which base set
16:11:00 <andythenorth> looks like a bug, but eh
16:11:25 <adf88> almost all dictricts are affected
16:12:35 <adf88> they are here: media/extra_grf/chars.png
16:12:48 <andythenorth> someone just drew it wrong?
16:13:39 <andythenorth> or are they supposed to be scaled down in code?
16:13:41 <LordAro> i'd imagine they date from a time when the boxes were fixed height
16:44:00 <frosch123> adf88: are you using opengfx or original?
16:44:55 <frosch123> i wondered whether it got broken when openttd.grf became the fallback for missing sprites, so it would mix standard letters from opengfx with diacritics from openttd.grf
16:45:30 <frosch123> so, most interesting is whehther it works when using spritefont and oriiginal baseset
16:47:36 <frosch123> or in other words: is one specific baseset broken, or does ottd mix basesets?
16:48:09 <adf88> I was referring to the original base set only
16:48:20 <adf88> what does it have to do with opengfx?
16:48:35 <frosch123> basesets define the glyphs for the sprite font
16:48:40 <frosch123> opengfx has its own
16:50:23 <adf88> anyway, original graphic set contains only some diacritics, and their sizes are fine
16:50:55 <frosch123> so, is it abuot not using the sprite font?
16:51:32 <frosch123> adf88: my problem is, that there are like 24 combinations for where those sprites originate from
16:51:48 <frosch123> and noone states which combination is broken
16:52:47 <adf88> heh, you are better since I have 0
16:53:10 <adf88> I have no clue where they came from
16:53:34 <frosch123> there are characters in baseset, in baseset extra grf, in openttd.grf and in non-sprite fonts
16:54:14 <frosch123> when i tried, i did not find a combination that fails
16:54:14 <Eddi|zuHause> non-sprite font, and that font's fallback font for when the character is not available...
16:55:11 <Eddi|zuHause> likely os-specific
16:55:24 <Eddi|zuHause> but i have no clue how stuff works
16:55:25 <adf88> I think we can just find some open font and make characters smaller in characters.png
16:55:43 <adf88> the fallback character from original TTD grafix are irrelevant
16:55:45 <frosch123> characters.png works just fine for me
16:55:59 <frosch123> .... why can noone just state which combination causes it... :(
16:55:59 <adf88> these characters are too big
16:56:29 <frosch123> they are not too big
16:58:50 <adf88> I mean that they should be smaller
16:58:58 <adf88> because original TTD letters are smaller
16:59:24 <frosch123> it looks just right
17:00:01 <frosch123> you and the reporter use some different fonts
17:00:05 <frosch123> maybe non-sprite fonts?
17:00:10 <frosch123> but they worked for me as well
17:00:24 <Eddi|zuHause> the screenshot in the bug report clearly uses a different font
17:00:54 <frosch123> adf88: my screenshot shows the sprites from chars.png, they look just right
17:01:01 <frosch123> i have no idea what the fs screenshot shows
17:01:37 <andythenorth> what font is used in the report?
17:01:55 <frosch123> because sprite font works
17:02:04 <Eddi|zuHause> maybe spanish has some characters that the sprite font misses, and thus switches automatically to non-sprite font?
17:02:28 <andythenorth> nor is it the chars in openttd.grf
17:02:32 <andythenorth> the T is highly distinctive
17:02:36 <frosch123> adf88: can you "killall -6 openttd"?
17:02:50 <frosch123> the crashdump should state exactly whether a sprite font is used or not
17:03:29 * andythenorth proposes just asking James in 6620, and waiting :P
17:03:40 <andythenorth> but maybe some of you like mystery puzzles more than me :)
17:04:16 <andythenorth> where do I get the sprite font?
17:04:27 <frosch123> you do not use a non-sprite font :p
17:04:37 <andythenorth> how do I do that?
17:04:42 <frosch123> and you do not use a weird language that forces a non-sprite font
17:05:02 <Eddi|zuHause> the least he should do is upload his openttd.cfg and possibly his savegame
17:05:45 * andythenorth bored, back to my BB game
17:10:10 <adf88> ok, it's strange, now I can't reproduce it
17:11:14 <adf88> in this FS, I think it may be some fallback indeed, not sure what caused this
17:11:43 <adf88> I'll try breaking some fonts :>
17:20:32 <adf88> I agree, on the screenshot the font is not the sprite font
17:20:43 <adf88> these character are not sprites anywhere
17:21:03 <adf88> but diacritics where not
17:21:23 <adf88> it might have to do something with font detection code
17:23:54 <frosch123> well, it also worked for me when switching to chinese
17:24:32 <frosch123> maybe it's the 9x build without icu :p
17:24:48 <frosch123> so many unknown variables :)
17:24:52 <Eddi|zuHause> maybe it's a newgrf which overrides the sprite font...
17:25:00 <Eddi|zuHause> but not all of it...
17:29:55 <andythenorth> the station icon for FIRS Explosives uses animated palette :P
17:30:02 <andythenorth> would be nice to see it
17:30:09 <frosch123> it's on position 5 of the list, so may be done in 25 weeks
17:30:42 <andythenorth> can’t Eddi|zuHause review it? :P
17:32:48 <adf88> looks interesting, i have no seee3 :)
17:32:56 <frosch123> andythenorth: eddi is busy figuring out whether neko and syl are the same person
17:40:41 <andythenorth> SYL needs some sig reduction
17:41:06 * andythenorth so much newgrf to do
17:41:13 <andythenorth> at least enough until 2020 or so
17:41:19 <andythenorth> FS holiday is over
17:41:39 <andythenorth> low score was 342, now people are adding more :P
17:47:06 <frosch123> andythenorth: can i add more wiki pages?
18:08:47 <Wolf01> You were looking for "ahead"?
18:09:04 <adf88> need to buy a new keyboard :P
18:09:26 <frosch123> cheaper than new hands
18:09:27 <adf88> keys are popping off, it's time
18:15:20 <andythenorth> frosch123: I just couldn’t get into the wiki game :P
18:17:55 <andythenorth> no obvious way to win
18:46:00 *** orudge` has joined #openttd
18:46:00 *** ChanServ sets mode: +o orudge`
19:06:07 *** Maraxus has joined #openttd
19:11:24 <andythenorth> it’s required at all 3 of the main iron/steel producing industries
19:11:32 <andythenorth> (indirectly in 2 of them as quicklime)
19:13:33 <Eddi|zuHause> Wolf01: just a coincidence that the comment is by "A. Wolf"? :p
19:13:42 <andythenorth> rest of the economy is remarkably balanced, stone is a bit sucky in comparison
19:19:08 <andythenorth> add more stone-producers? o_O :P
19:26:59 <andythenorth> limestone isn’t sea-dredged :|
19:27:06 <andythenorth> there are no sea industries in this economy
19:41:48 <Alberth> I didn't have lack of stones
19:42:01 <Alberth> you can't just give each industry a little less?
19:42:16 <Alberth> gung ho only asks that you deliver some
19:43:00 <andythenorth> I wonder if the problem is actually that it’s a bit boring
19:43:17 <andythenorth> three furnace industries, all need a stone derivative
20:04:25 <Eddi|zuHause> i was in the dolomites once
20:09:49 *** SimYouLater has joined #openttd
20:10:31 <ic111> ... but not implementable using the slope concept of OpenTTD^^
20:11:18 <SimYouLater> Is there a reason I would suddenly be logged off while using the forums? I'm worried someone may have gotten into my passwords when my computer got infected, but it might also be nothing since I've been logged in with "remember me" for awhile.
20:11:48 <SimYouLater> I only have had time to change my important passwords.
20:12:31 <andythenorth> probably can’t add magnesite
20:12:48 <andythenorth> already got manganese, too confusing, at least in English
20:42:25 *** quiznilo has joined #openttd
20:48:38 *** tycoondemon has joined #openttd
20:49:58 *** orudge` has joined #openttd
20:49:58 *** ChanServ sets mode: +o orudge`
20:53:21 <Eddi|zuHause> make an economy where all cargos start with M
20:55:26 <Eddi|zuHause> oha, thimbleweed park is on sale
20:56:45 <Eddi|zuHause> well, it's a fairly new game
20:57:24 <milek7_> good game, but disappointing ending
20:57:43 <Eddi|zuHause> i know nothing about the ending
20:58:11 <andythenorth> Magnesium, Metal, Manganese, Milk, Microchips
20:59:11 <andythenorth> 32 random M from dictionary?
20:59:16 <Eddi|zuHause> Military Supplies
21:00:48 <Eddi|zuHause> anyway, i have a headache that has been getting worse since noon :/
21:01:10 <Eddi|zuHause> and i didn't get any further in deponia
21:04:55 <andythenorth> swapped a newgrf cargo on a running game :P
21:26:40 *** FLHerne has joined #openttd
22:27:30 *** sim-al2 has joined #openttd
22:38:52 *** chomwitt has joined #openttd
23:22:25 <Wolf01> Mmmh, maybe it's time to go to bed early
23:50:08 *** HerzogDeXtEr1 has joined #openttd
continue to next day ⏵
