vehicle replacer, experience, usage behaviour

[sdog]

I'll briefly write about the cases where i used the replacer and where not in my last game. Then i'll give some thoughts.

usage narratives

Roughly my methods for replacing vehicles with newer can be put in three categories: manual, line based, total replacer.
In the early game i replaced only manually, sending out new convoys from a single, central depot on the map, and started to withdraw convoys as the new convoys arrived at the first stop of their line. It was necessary to have a gap-less service. The distances to the depot were quite large. Vehicles in question were petrol-electric buses by the first Letter-Type models.

The next method, after having a larger network and more income is a line based approach: Paused game, withdraw line, buying new vehicles in the d
epot, changing schedules to have them drive to different destinations and starting all (for buses) or starting them one after another to the same destination for trains.

I mainly use the vehicle replacer only to replace ALL vehicles of one type, regardless of line. Also i'm doing this in the developed game only for buses. In some cases i replace some key lines with the new model somewhat earlier, in the line based method above. The intention is to reduce the impact of dropping passenger numbers somewhat and more importantly not to overcrowd the key stations too much. In the recent game this applied only to the lines in Shrewsbury centre (11 and 12)

There are three exceptions where i used the replacer for trains. Replacing Auto train engines with more powerful Pannier Tank engines and extending them to two carriages. Also for all lines. It caused major interuption of my network, but it was acceptable due to the sheer number of lines. The short auto-trains also are less likely to cause complete clogging, when they used exotic ways to depots.

In one instance i replaced a large part of the trains on the express route Shropshire - Sheffield. Main reason was in that case that i confused some carriages and created a consist with too many dining cars. Since they are nearly identical i used the replacer to filter them out. The first disadvantage was that the replacer still does not recognize some trains with reversed consist as the same. Second problem was extreme congestion with deadlocks, as the trains took some very creative ways to nearest depots. Thereby geting on ways not appropriate for such long trains (7 tiles). I abandoned that project and gave the whole line a "withdraw all" order, deleting some blocked trains directly, and sending out new trains for the whole line from one depot.

I would have used it for some selected cases of the local trains, with medium sized engines and LSWR Compartment carriages. However the replacer can't be used for obsolete vehicles. As the depot where the replacing should take place is not defined, it doesn't matter that there are already such carriages available there. So i usually replaced with "withdraw all" and creating new trains in a depot. The 'copy convoy' action works also on obsolete, not directly obtainable carriages, so it can be exploited for this. The replacer can be used however, if the consist is in a form legal in the depot: in other words the break carriage is at the end, then removeing the engine and getting a new one with 'put in front' works. This has to be repeated several times however, after trains changed their consist at terminals. I used the replacer in this way to weed out the remaining M7 engines, after an apropriate BR engine below 65 t was introduced (4MF?).


Some thought on the replacer (quite opinionated)
Firstly, the replacer is absolutely priceless for one task, replacing buses, since they are in a often in lines with only one or two vehicles. Doing this in any other fashion is more than tedious.

For trains the system is more than tedious, due to more complex consists, higher dependence of the network on trains and congestion problems. For the much smaller number of trains, and more diverse types of trains (i only use two bus types at a time, local and travel bus) the benefit of the replacer is also much smaller. The improvement of train engines is also more incremental than that of buses up to the 1960s. It is desirable to replace buses as soon as a much better model is available. Trains can run considerably longer.

There was not one instance where i needed the more complex features of the replacer. Those would be skipping vehicles, retaining the vehicles in the depot, using vehicles already in the depot etc. Use home depot also was not quite useful to me, since this was not necessarily the closest. Mark for replacement could have ameliorated some of the problems i had before, but it is much more hassle than the other methods. I think i've tried all options already, so the reason for not using them is likely not unfamiliarity.

A small thing but a bit annoying is the default value for vehicles to replace is 1, and not the numbers of total vehicles. Also the arrows next to this field allow to set a range of 1 to 999, It should go from 1 to max vehicles in selection. (This number is displayed somewhere in the replacer dialogue already.) Inserting a number above 999 will highlight it read, the replacer still works though. As you know entering numbers is always a bit finicky in simutrans.

In my opinion, quite a lot of the features are there because you want to allow the 'cascading' of vehicles from one high profit line to an older line. I think, however, doing it this way contradicts a bit your position on micro management:

[...] generally, the rule is that, if players can gain an advantage by micro-managing, it disadvantages the players who do not micromanage, and makes the game less fun for them, and micromanaging is itself not very fun, so it makes the game less fun for everyone [...]

(link) This quite convincing me, and look a bit differently on some features in experimental (and simutrans in general).*

The cascading can be more easily achieved by allowing obsolete vehicles to be bought again. Withdrawing the old one credits their rest value. The difference can be just considered the cost of overhaul. If you actually install any of the overhaul concepts discussed in the other thread, this becomes even more desirable to do this, as the odometer would also be reset.

My suggestion is clean up the gui massively, drop most functionality (from the gui only at first) and see if there are complaints or requests for it. If not clean up the code, if yes you can decide what to keep. It should be enough to have replacement for single vehicle, all in line, all in game. Doing this always for every vehicle in the selected group. For all other things the game provides already the means to do so by other means.
edit: this was not quite appropriate to write, i just got too much involved in my own things while writing.

Oh, there is another reason that lets me avoid the replacer if i can do so easily, the convoy retains it's old name. This is particulary annoying with pak britain, where most buses look quite similar (doubledecked and red). I can't keep them apart at all. Perhaps check if the convoy still has the default name, and if it is so, have it renamed. If it is not the default name, the user changed it and it should be kept.

* I think i made it quite clear in other threads, that i think depots are completely useless altogether. Muchmore they are one massive source of micromanagement and also an inconsistency with the level of abstraction. An idea moblet brought up.

[mopoona]

You could make a depot menu instead of a building. All vehicles would have to be thrown onto the right track. Kind of the way you do with station placement. This may be an idea worth to be discussed.

[Banksie_82]

I think the replacer is great, I don't like to micromanage too much so this is a huge benefit to the way I like to play.

Unlike Sdog, I do use the "skip" and "sell" features within the window. Although at first it was confusing, the way to specify how many convoys to replace/sell/skip is also very useful. I would rarely use any of the other features however.

For example, say I want to replace all the buses (I usually only use one or two types) when a better one becomes available. I will select one of the buses, click replace, chose which bus to replace it with, click "Replace All", then chose for example to replace 1, sell 1 and skip 2. This sells a quarter of the fleet that I now no longer need due to increased capacity. Calls a quarter of the convoys back to the depot to be replaced and leaves half of the convoys running so the stations don't get too overcrowded. Once all of those that are to be replace are done, I'll then replace half of the remaining old convoys, and finally repeat for the last group.

I do the same for rail and trams (in fact all types of transport, often only 1 line at a time just to make sure it's all working smoothly). I have got into the habit now of setting up my network so that convoys don't block each other and can be replaced smoothly. This sometimes means setting up a bit of track that isn't used for the normal running of the lines. I find this quite realistic and I think it adds another dimension to the game.

One complaint however, when you try to replace more than 1 convoy the cost to do so becomes 0.00. Is this a bug, or is it just too hard to calculate?

In summary, after all the improvements that make passengers intelligently chose which route to take to their destination, this is probably the best feature of ST-Exp.

[jamespetts]

Thank you all for your helpful feedback; that is most useful. Apologies for not having replied sooner: I have been rather preoccupied in setting up my new computer.

One or two brief comments (as I do not currently have a great deal of time): the suggestions to abolish depots entirely would involve a very substantial change to the way in which the game works; there are a great many other outstanding issues that need to be resolved before such a substantial change can be contemplated (balancing features as discussed by Moblet, bug fixing, an industry re-linker and pakset work), and that is without consideration of whether such a change is desirable.

However, there are a number of more specific points that might well need addressing, including:

(1) the inability to use the replacer with stored vehicles no longer in production*;
(2) the incorrect calculation of the cost when replacing more than one convoy is selected (this indeed is a bug);
(3) the failure to rename convoys that have default names (I had, in fact, coded this feature, but evidently it no longer works: I shall have to look into this);
(4) automatically de-reversing the convoy in the replacer window and recognising as identical all convoys with the same vehicles in any order;
(5) thoughts about scheduling options to minimise the disruption caused by a replace operation (by, for example, automatically staggering the replacement over a period of time; perhaps by a setting to replace only X number of convoys at a time, not starting to withdraw more than the first X until the first X have finished being replaced).

* Allowing the purchase of out of production vehicles can be enabled at pakset level, but I'd rather not enable this for purchasing new vehicles, since this may give rise to economic distortions in cases where the price of a similar sort of vehicle is intended to increase. An option might be to list all out of production vehicles that are actually available in other depots, then to check whether there are enough of them in those depots, then to make sure that the convoy(s) get sent there, but this might take some time to implement.

If any of the developers would like to speed things up by coding any of these features, I should be most grateful!

[ӔO]

one feature I would like in the replacer, is "send back to last station convoy came from".

Massive traffic jams can occur when the convoy has to cross over a choose platform signal. For practicality's sake, I don't use choose-able platforms at all stations. The choose platform will normally be at the end of the line, or at a hub. When the train reaches the end of the line and is sent off to the depot, which can be on the other side of the station to keep the layout simple and short, the replaced convoy will cross a choose platform signal for the next station in line, and not the end terminal. This means the terminal station can become locked when all platforms fill up with trains waiting to be replaced and the replaced convoy is trying to find a route to the next station, through the waiting trains.

It's easier to replace all convoys if they are enclosed in their own loop, but it's just impractical to do that sometimes, because the station will be 16 tiles wide when 8 lines are connected to the station. This is partially solvable by stacking.


Sometimes I will use the skip and sell feature.
For skip, it's over the entire network, when I don't want all road vehicles rushing to the depot at once.
For sell, it's on a line by line basis, where sometimes, the replacement will prove to be too good at meeting capacity demands due to the speed and capacity boost.

[jamespetts]

Looking into this, I have fixed the bug at no. (2) above, but cannot reproduce the issue at (3) above. The other issues I shall look into when I have more time.