Falmouth Branch - a heretic's view

[AndiS]

I don't have the route, but since it is a general discussion of frequently seen issues anyway, I must say that while I share the theoretical criticism, I see these points also as simply sobbering regarding route building in our days. In the old days, you were happy if you PC would display some approximation at an acceptable framerate. Today, you PC would be up to it, but route building effort is multiplied by this.

Embankments. Long ago, Adam stated that for precise embankments you simply need to use lofts. They are intended for that purpose, the 8m terrain grid is not. This position sounds very reasonable to me, since you need to model the side path (cess) and a decent shape of the slope.

My personal favorite disturbance is the amount of ballast that is wasted 5 m besides the track. If in fact they would ever do that, it would be overgrown and thus invisible.

Incisions. Lofts would be even more important here, since the slope is much more prominently noticed by the driver. Also, the water drains on the side form an important part of the overall structure. However, to form the top of the incision nicely, you would need a second loft for that, and this seconds loft (some A-profile) whould show Z-fighting at gradient breaks (which are unavoidable since all incisions go through some sort of hill). Of course, you could hide the bad spots using lots of vegetation but at least for UK and earlier periods, it is not as prototypical as for most of Europe today, where jungle hides everything around the line.

On a side note, with vegetation and earthworks enclosing the train path as it does in many places, helicopter view would be a quite different experience.

Bridges. For years, I dream of creating a bridge building assistant in Blender. The fact that the professionals don't have it in 3DSMax seems to indicate that most customers have no idea about bridges. Actually, in many cases, you don't need a clever software assistant, you just sweeze the object a bit. But somehow, bridges which are high enough but not higher, and which are supported where the girder work demands it, do not seem to be on the agenda.

Likewise, if you run the route on detail level 10, you are a happy customer. If your PC does not permit that, you will not be a happy customer anyway, because you cannot appreciate their complete effort, so the shades of your unhappiness are a minor concern.

On a side note, I must say that I myself am cautious when judging bridge shapes since there are many early concrete bridges with a stone or brick coating, and they were considered elegant around the time when railways were built.

Manual switches. Their reason is our favourite two-letter word, AI. The dispatcher is said to be a pain when it comes to reversing and plotting extensive shunting movements. Also, part of the community loves shunting under their own responsibility, setting the path to the various tracks themselves. This leads to the requirement to have manuals switches in the locations concerned. On many smaller routes, mainline track is used in the course of shunting movements.

However, the consequences need not be so draconian. But since this is a game, it is not easy to devise a solution. An easy solution would be to replace the lever shape by a switch stand which is more an indicator than a lever, or exclusively an indicator. However, this is not prototypical either for many places, because when the switches were set and locked remotely, the driver could rely on their position, so an indication was not needed, at least for the passing trains. But often, some indicator was still used, and it might be possible to find a UK example which suits many cases, better than the currently used lever.

Like several others already suggested, you can move the switch levels to the side, which permits their grouping into a ground frame. Of course, the design would be a different one. But the main problem is that levers of prototypical size at prototypical spacing are hard to click at from some distance. On the other hand, it would cut some flying during shunting if you could just fly close to a single ground frame and set several switches from there.

Signalling. I guess a few new default objects are all that is needed to make the given functionality look much better.

a) A signal with a big (normal) and small arm. It would work exactly like the existing bracket, but look less out of place in many locations.

b) Two, three, maybe four disks stacked, with the corresponding track link count.

c) A minor modification of the script logic to make it ignore passing trains at link 0 when no path is set. This permits the usage of two different signals side by side, for different branches of a junction (or station entry etc.). It is just a few words in the code. But I have not tested AI behaviour in such a situation. But then again, AI is not as bad as its reputation on a properly signalled route.

d) Alternatively, a group of 1E signals which do not clear for the yard entry link. They would be complemented by disks which do not deal with track occupation, clearing the block in approach would be performed by the 1E signal. All you need is a few words of new code, again.

e) Maybe a revision/completion of the shunting signal class system. Some are just switch indicators, not taking part in the general train announcement scheme. Some do take part in it. Some only announce, but clear without announcement (exit from "yard area"). Some clear on announcement but forget the train as it passes (but report it back), they are what you need at "yard entries". Ideally, you would need all four classes, in all reasonable link numbers, and all shapes (disk and light signal, UK, German, US?). Implementing (d) cuts the requirement list under (e) significantly.

Besides such a wish list, there are also a few simple facts to use already now. AI only works well on fully signalled track. However, if the signal is not important otherwise, you could just sink it under ground. But I don't think there is such a signal. What people may wish for are shunting signals which are seen as block separator by the AI.

However, for bigger yard areas, it can take some serious planning exercise (in the prototype, too) to devise schemes under which more than one engine can work safely and without too much mutual blocking. And with most part of the game not knowing what is to come in the next minute, I cannot see it coordinate complex shunting movements (which include lots of changes of direction) of several engines near each other, simulaneously.

But then again, with the multiplayer plans just reconfirmed, we might see a dispatcher who can adapt his plans to the course of events, one day. Unless they replace him by one of the players, which leads us far off topic, so let's speculate about that elsewhere.

[FoggyMorning]

Just a further thought regarding the use of main aspect signals on sidings; in the prototype presumably some of these would be protected by ex-GW "ring" shunting signals, for which there is no really appropriate substitute in the default assets

[Kromaatikse]

The ringed signals used by the GWR had exactly the same meaning as normal Home signals, the ring only helped identification of the slow-line signals.

Dwarf arms and discs with various markings were used for shunting, though.

[FoggyMorning]

My understanding is that ring signals were used to indicate exit from sidings or yards onto main running lines, and dwarf arms to indicate entry. Typically this was an inconsistent practice and ground discs could also be used to indicate the same

[stuart666]

My understanding is that ring signals were used to indicate exit from sidings or yards onto main running lines, and dwarf arms to indicate entry. Typically this was an inconsistent practice and ground discs could also be used to indicate the same

Thats right. In most cases inside the yard, the points seem to have been indicated by dummies, but usual practice running from a running line into a yard or siding seems to have been a siding arm, and running out seems to have been a disc. Im sure there were exceptions, but that seems to be the usual setup I see in station plans.

[freightonrails]

I would like to buy this addon, it would be my first addon purchase. The pc I use for railworks does not have an internet connection can I still buy an addon on another machine and copy it across ? Thanks

[markpullinger]

Begs the question as to how you got Railworks on that computer without an internet connection in the first place! It comes down via steam so will have to be installed directly from the internet. You might have to take your machine to where you can get connected.

[Kromaatikse]

I think Railworks2 is available on disc and will install from that independently - but without Internet you won't get any updates or DLC.

Steam will work on a pretty weak conection though - you will just have to be very patient about waiting for downloads.

[Darpor]

I think Railworks2 is available on disc and will install from that independently - but without Internet you won't get any updates or DLC.

Steam connection is still required as stated on the disc version.

[Blackmonk]

. Still, if it's any consolation, MSTS didn't do it any better and Trainz effectively they're all hand points!

Good thread.

I've seen this written several times and frankly it is not correct. There are several rules in Trainz which allow paths to be set and points to be locked in position in a session. There is also an advanced scenario creation tool called SCS2006, which gives total contol over the game elements for whatever you want to do.

In terms of the OPs points, I find the the abrupt gradient changes in some routes one of the things I enjoy least in RW.

regards
Ed

[bigvern]

I've seen this written several times and frankly it is not correct.

I hear what you say but the simple fact is, out the box and without any scripting or enhancements Trainz relies on the user to find their own path to the destination - if it varies from that pre-set in Surveyor. Even on a simple single track route, if you set up the points approaching crossing loops to cater for arriving trains, on departure you need to manually throw the point to get a green signal (assuming no opposing train is approaching). It wouldn't be so bad if you could "save" the actual player route with each session but you can't, only as the switches lie in Surveyor level.

I'm aware that you can give trains routing instructions but player control of a train over-rides these and, for example, arriving at Kings Cross throat with the default route set for Platform 2 but you want P10 means faffing about throwing various levers until you get the correct route set. As regards the third party advanced scripts, obviously these have been developed because the basic despatcher is so lacking. Whether these are included by default or must these be obtained from the DLS, it's a fair bet that 95% of Trainz users either aren't aware of the existence or how to use the application. Looks like I have a learning need on that one. While I believe a version for the later editions is in production, SCS2006 only works on earlier versions not 2009/2010 and again is presumably a fairly specialised scripting application.

Now please don't get me wrong. I like Trainz alot for all sorts of reasons and (becoming fed up once more with Railworks myself) it's looking more and more likely I'll be building my next route in TS2009, but the lack of a core autonomous routing despatcher always has been and still is one of its weak points.

[freightonrails]

Thanks for your help. I should have said that ONCE the pc did go on line to start Railworks but its too valuable to back on the internet. My Laptop can, so I thought it may be possible to buy it put it on the laptop and copy it to the main pc.

[Blackmonk]

I've seen this written several times and frankly it is not correct.

I hear what you say but the simple fact is, out the box and without any scripting or enhancements Trainz relies on the user to find their own path to the destination - if it varies from that pre-set in Surveyor. Even on a simple single track route, if you set up the points approaching crossing loops to cater for arriving trains, on departure you need to manually throw the point to get a green signal (assuming no opposing train is approaching). It wouldn't be so bad if you could "save" the actual player route with each session but you can't, only as the switches lie in Surveyor level.

I'm aware that you can give trains routing instructions but player control of a train over-rides these and, for example, arriving at Kings Cross throat with the default route set for Platform 2 but you want P10 means faffing about throwing various levers until you get the correct route set. As regards the third party advanced scripts, obviously these have been developed because the basic despatcher is so lacking. Whether these are included by default or must these be obtained from the DLS, it's a fair bet that 95% of Trainz users either aren't aware of the existence or how to use the application. Looks like I have a learning need on that one. While I believe a version for the later editions is in production, SCS2006 only works on earlier versions not 2009/2010 and again is presumably a fairly specialised scripting application.

Now please don't get me wrong. I like Trainz alot for all sorts of reasons and (becoming fed up once more with Railworks myself) it's looking more and more likely I'll be building my next route in TS2009, but the lack of a core autonomous routing despatcher always has been and still is one of its weak points.

To avoid going way off topic, I'll just say this :
- there are ways round the KX situation you mention e.g using a Path Rule,
- SCS2006 does work in TS2010, but an updated version is being worked on.
- SCS2006 outputs a session environment which a user can play - very similar to an activity in MSTS/RS/RW.

regards
Ed