Return of the Backdated North London Line

[J1o2h3n4]

Not sure if anyone else had this but in a scenario where I left Broad St in an EMU, another EMU was entering the station on an adjoining line there was insufficient clearance between the tracks and I collided

Yes had the same thing happen to me.

Regards JohnT

[J1o2h3n4]

allowing the scenario "03 - One For The Scrap Heap" to be completed. was the original problem a Red signal, which did not respond to a TAB request nor seem to have any reason to be set at Red. I ignored it eventually and then completed the scenario.

Regards JohnT

[JustRight]

Great detective work guys. I for one would never have thought of changing the Loco type. I'll give that a go myself and see how I get on.
I'm sorry that I missed the conversation, but being on the other side of the world doesn't help.

Thanks to both CeeGee for the scenario's and Xguerra, for a fabulously rendered version of the NLL.

Cheers,
Trevor

I have this evening modified CeeGee's scenario, replacing the Kuju Class 37 with the DTG BRBluePack01 Class 37. With no other changes, using the newer version of this loco allowed the scenario to run correctly.
To paraphrase CeeGee, I'm gobsmacked. Never seen a scenario affected like this, in this way, before.

Oh well, we live and learn.

Thanks guys.
Trevor

[CeeGee]

I have been through all of my NNL(Backdated) scenarios and removed the Kuju Class37 (just to be on the safe side).

They are available here:
http://www.wotawallysbits.co.uk/my-scenarios

[JustRight]

I have been through all of my NNL(Backdated) scenarios and removed the Kuju Class37 (just to be on the safe side).

They are available here:
http://www.wotawallysbits.co.uk/my-scenarios

That's great mate.

Thanks very much!

[CeeGee]

If you have the original scenarios then you can swap the Kuju Class37 in RW_Tools to any other model you have. The easiest way is in the RW_Tools Scenario Editor find a Class37 (there are plenty of consists titled 37 or CL37), select the item you wish to swap it with and then Click on "Swap in ALL scenarios for this route"

[gptech]

Before panic really sets in and the whole world becomes convinced that the Kuju default class 37 is *evil* could somebody see what happens if you swap it with either another livery version of it, or even with itself. Lets ascertain, if we can, that we aren't looking at an edit to the scenario being the trigger for the game to recalculate the pathing information resulting in a happy ending.

[CeeGee]

I had already changed the Player Class37 from "Special" to "Express Freight". I had already tried altering the timing of the Player Class37 by a small amount - and done the same for the Class37 AI consists that used the crossing.

[gptech]

Yes, the crossing....apologies for going back a bit, and also if I've missed an answer, but do these *problem AI services want to use the *right* crossing.

Has anybody tried replicating Xavier's test scenario but using, for example, the default Kuju 47 from the outset?
Whilst we know an edit to a scenario works we haven't yet determined fully just what is or isn't affected in a brand new scenario with other stock.

[CeeGee]

There is a very simple way to find out - make a scenario.

[gptech]

There is a very simple way to find out - make a scenario.

Very, very, simple, and if you read again what I've been doing you'll see that I have already made more than one scenario to test with. Once I've had my tea--I've only just got home from work after a 12 hour shift--I'll cobble something further together.

If you also read again you'll hopefully understand that the question wasn't "which crossing do the trains use?" but "which of the two available paths would be the *correct* (as in most prototypical) one?" I even re-posted davep's picture, not the best in the world I'll admit but it does more or less show them both on the 2D map to make it as very, very simple as possible to give a useful answer.

[davep]

Gary, the westbound coal train should use the first crossover (the one circled) IMHO.

[CeeGee]

The correct routing is through Camden Road P1 which is where it was sent originally. Which is where it normally goes. It reaches the junction a few minutes after the player train. Davep's picture is of a broken scenario where he stopped at the signal for five minutes. Even in the broken scenario if you TAB straight through the signal ignoring the "denied" the AI service will take the correct pathing.
To suggest that someone should make a minor change is insulting to me by inferring I did nothing to find the problem. I tried many things and made a few suggestions to help actually run the scenario as it was. When my workaround was applied then players were completing the scenario. There were also those that cloned the scenario themselves and made small changes, trying to help and find answers.

[gptech]

The correct routing is through Camden Road P1 which is where it was sent originally

Yes, but how did it get to Camden Road P1, and how should it have got there is what I'm trying to ascertain---there are 2 possible junctions, so which one should it use?...we know it uses the second (from the viewpoint of it's line of travel) but as Dave has said (and I agree) it does seem to make more sense for it to use the first one, which I circled on the re-posted photo.

It reaches the junction a few minutes after the player train.

Yes, but we've moved on from your scenario and trying to replicate/corroborate/refute Xaviers test and findings.

To suggest that someone should make a minor change is insulting to me by inferring I did nothing to find the problem.

Where did I suggest a minor change to your scenario? I posed the question "Isn't the easiest solution in cases like this to use scenario specific markers to *force* routing?" as a general one, in the part of the discussion about Xavier's test scenario but NOT about specifically about anything you've done.

I tried many things and made a few suggestions to help actually run the scenario as it was

Which it didn't need in the version of the route it was written for, which brings us back to finding out just exactly what has/had changed and why substituting the loco in the updated route worked. Blaming the class 37 in entirety and drawing a line under it is flawed methodology---more on this after the pictures and test findings.

Replicating as far as possible Xaviers test with slightly different stock---same vintage (game wise) so just like the Kuju 37 there's no advanced scripting to influence things. (The scenario was created, saved, and the game exited completely and only after a restart was the scenario played)

Player train starts at Camden Junction



AI service at Caledonian Road



Signal NL1205 is green,



(photo taken from second man's side so you can just make out the red glow from signal NL1213)

So...you've a clear signal (should be yellow with a danger aspect ahead) but you carry on to stop at NL1213. But....the points are set against you before you reach the signal



and you find the AI service bearing down on you.

Not a class 37 in sight, the player train should have been held at NL1205 or the AI held at 3103(?) to allow the player train through. This obviously could do with testing in the original V2 edition, if it works as it should in that then the most obvious cause of this particular issue is the subsequent track edit, and NOT the Kuju class 37. That's not saying that it isn't the class 37 (even though I'm convinced of that) so to try and clear that up let's see what happens when following Xavier's test a little further---swap one of the (now also evil) class 66's for something else.... (done with the in-game editor, saving and exiting the game completely before reloading)



Yes, a *broken* default Kuju class 37. Signal NL1205 this time showing green,
Signal NL1213 now however is clear, with the feather indicating the correct route.



and we find the AI held at 3103 ...



Now do you see why I'm sceptical about the claim that the class 37 is the root cause?

To theorise a bit:

changing the service class didn't make things right, though you'd certainly expect that to work--conventional thinking has it that it should cause a complete re-work of the pathing, but it would seem that isn't the case, at least not in this instance, so is that change merely amending the data about the service, and it needs a change to the content of the service to force such a low level and fundamental rework?
That goes some way to explaining why I had no issues with "Freight to Stratford" as because of my particular *madness* and love of having stock in non-supplied locations I had to do the editing with RW Tools before playing.

I'm sorry if that looks like too much of a minor change, and feel free to disregard my findings as irrelevant but please don't try and tell me that the class 37 is the root cause of the issues. It matters not one bit if the signalling/pathing *fault* is down to the route, the scenario, the stock or a combination of any or all of those as long as testing is done methodically, and every possible (no matter how unlikely) avenue explored rather than latching on to the first possibility as being THE reason.

The next test, and it won't be tonight 'cos I've a cold woman and a warm beer ( wrong way round, I know) waiting for me, will be to revert back to the *broken* original class 66 scenario and see if changing just a wagon (the content of the service in an even more minor way) has the same effect.

Have a think, even have a go yourselves and see if you get the same results, or even completely different ones and we'll eventually work out just what's going on.

[xguerra]

So...you've a clear signal (should be yellow with a danger aspect ahead) but you carry on to stop at NL1213. But....the points are set against you before you reach the signal

Now that's strange. I do not have that set of points................that changes things a little. I'm racking my brain now to assert when I put those points in and when I took them out.....

And my summation originally that the issue originally was not that the Kuju 37 was used at all, but both trains were a Kuju 37. Hence why I said "if one of the two offending services in Colin's scenario were changed from a class 37 to a class 47 or another loco, it would run fine". Effectively, this is proved by your demonstration above as one of the two trains is a 37 and all appears to run smoothly.