RSC 2-8-0 CONSOLIDATION

[mrennie]

I've added a few lines to the HUD version's script so that it forces the exhaust injector to operate when the loco is an AI and the boiler water level dips below 80%. Hopefully that'll fix the problem.

Good news.
Lets hope a possible core bug doesn't override the script instructions.
Just to clear out:
Is this implemented in the second update for the AI version, or is it for the next?
I mention it because the consist I built (and failed with the water error) had the HUD for both locos in the double header and I only used the "AI helper" version in the rear (3 consolidations in total).

It'll be in the Steam update that I'm finishing now. I'm also having to add more stuff to the script so that the computation of the smoke colour based on the mixture of air, draft, coal and volatiles, works for the AI locos.

[mrennie]

I'm testing a heavy consist on Marias Pass, double-headed with an Advanced (player-driven) with a HUD-enabled (AI) behind it and an AI Helper at the rear. So far so good. I've saved it just after leaving Whitefish and I'll continue running it tomorrow until the end (it'll take several hours) to see if I can reach Shelby without a hitch.

The smoke from a double-header looks nice:

2015-10-15_00001.jpg

I've also modified the engine blueprints and scripts so that if the windows, doors or roof vent in the player loco are opened/closed, the AI locos in the consist do the same (I realised it looked strange for the player to have his windows open but the AI locos to have theirs closed).

[ozinoz]

The smoke from a double-header looks nice:

Master of the under-statement

[slick204]

Loving the updated Connie! Thanks again for all the improvements.

Working on the RW&A. Love that smoke.

ConnieRWA.jpg

[jalsina]

Mike,
The smoke looks great in Marias pic.
What is the consist composition in that double headed?

Something I wanted to mention about that crash due to AI water depletion in the uphill Horseshoe run.
I had trouble in some places during the uphill run and almost stalled the train in some moment but never went to zero mph. In those instances the second loco released cocks steam every time the train crossed the speed of 3.3 mph, either decreasing or increasing. The same happened with the rear AI helper. Imagine a speed variation like this:
8, 6, 4, 3.3, 2, 3.3, 5, 3.3, 2,......Etc...
Every time the train ran at exactly 3.3 mph the AI Connie opened the cocks.

[mrennie]

Mike,
The smoke looks great in Marias pic.
What is the consist composition in that double headed?

I just put a lot of stuff together at random. I'll check to see the weight.

EDIT: 1,124 Imperial tons (1,259 US tons). I checked it by printing out the result of Call("GetConsistTotalMass") and noticed that once the train got underway, the mass continually decreased, which must be the coal and water being consumed.

Something I wanted to mention about that crash due to AI water depletion in the uphill Horseshoe run.
I had trouble in some places during the uphill run and almost stalled the train in some moment but never went to zero mph. In those instances the second loco released cocks steam every time the train crossed the speed of 3.3 mph, either decreasing or increasing. The same happened with the rear AI helper. Imagine a speed variation like this:
8, 6, 4, 3.3, 2, 3.3, 5, 3.3, 2,......Etc...
Every time the train ran at exactly 3.3 mph the AI Connie opened the cocks.

When used as AI, the locos are scripted to open the cylinder cocks when moving at no more than 3.3mph. That's intended for when they're in an AI-only consist and moving off from a standstill. I'll modify the script so that when the AI loco is in a player consist, it does what the player does.

[jalsina]

I just put a lot of stuff together at random. I'll check to see the weight.

EDIT: 1,124 Imperial tons (1,259 US tons). I checked it by printing out the result of Call("GetConsistTotalMass") and noticed that once the train got underway, the mass continually decreased, which must be the coal and water being consumed.

It is about the same weight of that consist I used in the HC. 16 mixed cars I believe.
Very interested in knowing the final result of that Marias Pass crossing. I had an extra helper behind.

How do you read the train mass underway, an scenario script?

[mrennie]

I just put a lot of stuff together at random. I'll check to see the weight.

EDIT: 1,124 Imperial tons (1,259 US tons). I checked it by printing out the result of Call("GetConsistTotalMass") and noticed that once the train got underway, the mass continually decreased, which must be the coal and water being consumed.

It is about the same weight of that consist I used in the HC. 16 mixed cars I believe.
Very interested in knowing the final result of that Marias Pass crossing. I had an extra helper behind.

How do you read the train mass underway, an scenario script?

I read the mass by putting a debug print statement in the engine script:

Code: Select all

SysCall("ScenarioManager:ShowInfoMessageExt","Debug",Call("GetConsistTotalMass"),0,0,0,0)

I've just been modifying the scripts and blueprints again to make sure that the AI helper mimics everything, including opening/closing doors and windows, smoke density and colour and the opening/closing of the cylinder drain cocks, as well as topping up the boiler. I'm going to start the run again, Whitefish to Shelby, and see how far I get. I'll try it once with an Advanced engine leading the consist and then with a HUD version - in each case the second engine will be a HUD and there'll be an AI Helper at the rear.

[mrennie]

It turns out that the AI locos can't even give their own script any information about the level of water in the boiler. Call ("*:GetControlValue", "WaterGauge",0) always returns 0, even at the start of the scenario. So there's nothing the script can do to counteract the core code "steam double-header" bug.

[jalsina]

It turns out that the AI locos can't even give their own script any information about the level of water in the boiler. Call ("*:GetControlValue", "WaterGauge",0) always returns 0, even at the start of the scenario. So there's nothing the script can do to counteract the core code "steam double-header" bug.

But yet your script for AI filling water when it goes below 80%, would or would not work (or at least implemented), given what you are mentioning in this post (even if the variable keeps reading 0)?

I hope DTG /Kuju are in knowledge of the bug which comes at least from TS2015, but probably is one of those core older codes from Kuju.
It seems there has been some problems in the newest European locomotives like the Peppercorn K1 and in Riviera50s.
I don´t remember well, but I think a similar bug was also present in MSTS.

I read the mass by putting a debug print statement in the engine script:

Code: Select all

SysCall("ScenarioManager:ShowInfoMessageExt","Debug",Call("GetConsistTotalMass"),0,0,0,0)

Once you have it implemented in the Consolidation scripting, how do you print it? Is it a key stroke that displays a screen reading?

The only way I know to get the mass of a train is using the Custom consist editor. I don´t even know if there is something to read it in the Build scenario editor.

[mrennie]

But yet your script for AI filling water when it goes below 80%, would or would not work (or at least implemented), given what you are mentioning in this post (even if the variable keeps reading 0)?

I hope DTG /Kuju are in knowledge of the bug which comes at least from TS2015, but probably is one of those core older codes from Kuju.
It seems there has been some problems in the newest European locomotives like the Peppercorn K1 and in Riviera50s.
I don´t remember well, but I think a similar bug was also present in MSTS.

It's one of those bugs that, I think, will never be fixed.

The problem with what I tried to do in the script is that I can't simply turn on the injector in the AI loco when the water level falls below, say, 80%, because there's no way of knowing the true level. So the script was keeping the injectors on all the time (I realised that when I noticed the constant sound of the injectors, coming from the AI loco). I don't even know if the injectors make any difference to the AI loco (as I can't read the water level in them). Basically, the problem with the AI locos is that they operate on very, very simplistic rules. Virtually every core code controller that you try to read from an AI loco comes back with either 0 or a null pointer. You can read the speed, but that's just about all you can read from the core code, for an AI.

Once you have it implemented in the Consolidation scripting, how do you print it? Is it a key stroke that displays a screen reading?

The only way I know to get the mass of a train is using the Custom consist editor. I don´t even know if there is something to read it in the Build scenario editor.

I put that line somewhere in the "Update" function in the engine's script, so that it gets called continually (roughly once every 10-20ms). The call to "ScenarioManager:ShowInfoMessageExt" makes the message appear in a dialogue box in the centre of the screen. I find that more useful than using a Print call that puts the text into one of the LogMate tabs, because I can watch values changing in real time, and it's not limited to control values (I can put anything I like in that message - even a simple "Hello" to know that I reached a particular line in a particular function in the script). There are other ways of debugging - another is to run TS in windowed mode with the "-ShowControlStateDialog" launch option, so that you can watch the values of all the controllers.

[jalsina]

The first part of your previous post doesn´t feel very promising
It is about of coupling a double headed and praying

[trev123]

I found an old post of mine posted in 2012 that I had forgotten about and said when double heading turn auto fireman on and the random running out of water error doesn't happen.

[jalsina]

I found an old post of mine posted in 2012 that I had forgotten about and said when double heading turn auto fireman on and the random running out of water error doesn't happen.

And apparently as mentioned in some posts, the F5 value for water level shows different than the HUD. But there we are reading the head unit and not the AI.
Using the auto fireman doesn't look too charming and definitively not for the Consolidation, considering the scripted fireman.

[mrennie]

Using the auto fireman doesn't look too charming and definitively not for the Consolidation, considering the scripted fireman.

There's no expert scripted fireman in the Connie (it's only in the FEF-3) and youcan use the game's auto-fireman with both versions of the loco (HUD and Advanced). What is scripted is the automatic opening and closing of the firebox doors according to the stoking rate.

That said, the game's auto fireman is rubbish. It's obvious that Kuju knew that their auto-fireman can't keep the boiler pressure up, because (and this is from my own observations of what happens in the game) they must have hard-coded it so that when the boiler pressure drops to 65% of the maximum, it artificially stops falling. There's really no need to use the auto-fireman, because in HUD mode it's really easy to fire anyway (just press the firebox doors button as needed to keep the level at 60% and the water button whenever it looks like the water is low - say less than 80%). Even in advanced mode, it's not that hard, as long as you remember to operate the injector water valve and lever in the correct order (and make sure the steam throttles to the injectors are open).