Forum Replies Created
-
Posts
-
Buffer stops
It would be nice if a buffer stop is placed automatically on every open track end. In reality there are no open tracks – and you could simply detect bad connected tracks visually.+1 for this. And in addition, same thing for the electrification would be great (something like an extra pylon and wire connecting to the ground or something; but not on the stations). It really would help with debugging electrification on routes. Of course, an overlay mode to colour electrification as when hovering on with the electrification tool would help even more. Just an extra checkbox to the view mode panel on the bottom right. Perhaps also an colour overlay with top speeds as well? But with a gradient of colours, not just green and red.
Recognize connected freight lines
Right now the production company produces only goods for direct connected lines. But I want to transport the goods to a freight station and reload them into another train to transport them to another city. That is not possible at the moment.
@The_President, are you really sure about this? By my observations, cargo works just as passengers; they find their way to a final destination, even using multiple lines. Most commonly I use trucks for the end delivery (from the cargo station to the city’s industrial areas), but I can’t believe using trucks or trains for the first leg wouldn’t work either. Usually the problem that the cargo won’t get produced and hauled onto a line is that there isn’t enough demand for it.@kimmaz, what is your game’s total population? The freezes really don’t start until well into the game. Depending on the processor power, I’d say between 10,000 and 20,000 of the total population. Probably a hefty cargo production as well. If you only have about 20 trains, I’d say your world population just isn’t large enough yet.
I now made some more tests with the clock. It seems that it’s not so much end-of-month freeze, but finishing up the calculations made every month, but started already in the beginning of the month. The month in my game always takes around 2 min 30 seconds. If I just let it run, the month itself takes around 1 min 5 s, and then freezes for 1 min 25 s. (Quadruple speed clocks to much the same; the month itself isn’t any faster.) If I pause it just before the end of the month for 1 min 30 s, the month ending takes no time at all. And ALSO if I pause it on the second day of the month for the 1 min 30 s, the rest of the month takes the normal about a minute and makes it to the next month instantly.
So, the monthly calculations start at the beginning of the month, and the game just waits for them to finish before letting the game run the next month. That’s why we won’t see any delay at all until the game has progressed quite a lot. It always has that about one minute (or at least 15 seconds on quadruple speed) to do what it is doing.
So the problem now just is that is once-a-month fine granularity enough for these, or should once-every-three-months suffice? Also, do the calculations really have the most optimal algorithm? Is it just plain-old Dijkstra or could there be an improved choice? This is the worst bottleneck of the game, and probably not even bound by processor power as much as the sheer algorithm complexity.
@krogoth_mk2, I’m not insinuating it’s a problem with the number of trains or vehicles. My original description is just a general description about the state of the game, not anything else. I think it’s much more a problem with the population, and perhaps with the network complexity. My hypothesis is that all the people agents in the game get their job, shopping, and leisure destinations recalculated each month. But as you suggested, it might also be city-growth-related, because as I described, I made quite a large overhaul in the central city, and that grew the end-of-month delays from around 20 seconds to a total minute. Maybe I’ll let the game run for a few years on its own and remeasure after the central city growth boom drops.
But for the numbers, as asked: I have now about 50 trains, 150 trams, and 150 road vehicles. My central city has a population of 2,000, its neighbouring cities (4) population of 1,000-1,500, and the rest of the connected cities (7) 500-1,000. Totalling about 15,000 people on the map.
Even though the maintenance fees also occur during the end of the month, it’s quite impossible that they would be the reason for this kind of lag. There’s not much to calculate, even if there were 50,000 vehicles. This is something completely different.
@kimmaz, had you read carefully, you’d have noticed I spoke about the freeze once a MONTH. Autosave happens, as you yourself said, once a YEAR. And a freeze much much worse than just a second. As I said, it takes a whole minute. Furthermore, modern autosaves take even less than a second, because they only need to pause the game for duplicating the game situation in-memory. After that, the disk writing can take its time while the game continues playing. But as I said, this is completely unrelated to autosave.
BTW, your list of coaches is missing Edo: http://fi.wikipedia.org/wiki/Edo_%28ohjausvaunu%29
It’s a must if you’re going with Ed coaches. There’s no contemporary Finnish railways without control cars. Just leave the Hr11 and other curiosities to be, but please make the Edo. 🙂
Looks great! Very lifelike and realistic. I’m a Finn, too, so I know. 🙂
May I suggest to you next to take under work the Sr2 and Sm5, because there are already models for Re460 and Norwegian FLIRT, so that it would only require redoing the textures (if the modders of those allow it, of course). Hopefully Sr2 with both the pre-2009 (red) as well as post-2009 (green) livery. Or if you’ll have to choose, only the red livery. Of course, the S2 needs Ed-series coaches to run with, with the new control car as well…
I have never made mixed freight-passenger trains, and probably never would. The thing is, cargo and passengers need to be delivered to different stations, so that would mean twice the stops between cities. Even on early trains, it would make them much slower. Also, the route of the cargo is usually just from the factory to the city, so that even the route would have zig-zag much between the cities. In my latest game I have made complex cargo network just for the fun of it. It probably doesn’t run even on profit. The cities just have a certain demand for cargo, and the nearest factory is good enough for them. The factory scales its production upwards until the demand has been met.
