Signalling and AI Traffic

[mbtech22]

I am currently working on some minor changes to the UK semaphore distant signal scripts and I could do with some input from route builders and users.

At a UK semaphore distant signal would you prefer that its aspect is displayed on the F3 and 2D maps and the AI traffic treat the signal as a block signal. Alternatively would you prefer it as now, with the signal invisible on the F3 and 2D maps, but with the AI traffic ignoring the signal and the signal blocks running from stop signal to stop signal at per the real thing.

Unfortunately these limitations appear built into the main programme and are flagged in the xml file and have nothing to do with the signal scripts.

Please let me know your constructive thoughts of which of the alternatives you would prefer for single arm distant, splitting distants and fixed distant signals.

This will enable me to consider how to set these options for the soon to be released next SR signalling kit.

Thanks
Mark Brinton

[Sly401]

Hi Mark

I am not a route builder but as a user...

Can you display the colour as Amber if the distant is on ? or are we limited to red and green?

If you can show as Amber it would be better to display ... if its going to show as red then it would be better ignored

Thanks for your work on this..

Sly

[mbtech22]

Hi Mark

I am not a route builder but as a user...

Can you display the colour as Amber if the distant is on ? or are we limited to red and green?

If you can show as Amber it would be better to display ... if its going to show as red then it would be better ignored

Thanks for your work on this..

Sly

The F3/2D display can be made to show a yellow or green for distant signals. However if we do this then you may find AI traffic appearing to follow you from the previous stop signal. The AI traffic will treat the distant signal as a stop signal. It should not cause any collisions though as the following AI train will not leave the preceding stop signal until you are passed the distant. The AI train will then proceed as far at the distant and stop if you have not cleared the next stop signal, irrespective of the aspect displayed by the distant.

Any more comments, please.

Mark Brinton

[AndiS]

I would not declare the distants stop signals. I would rather have the shunting signals disappear from the map. Of course, showing them in another colour or another symbol would be better. But then again with all the disks stacked on each other, you would not see it most of the time.

Btw., by the same token the trick with implementing each signal doll as a separate signal and have lots of link 0s collocated next to the bracket post makes the map view a bit useless, regarding signal state, doesn't it. I mean, you only see the colour of the topmost disk, all the others hardly peek out a bit from under the top one.

I found that setting the colour to -1 or 3 or some other "forbidden" value paints the signal black, or it does not paint the signal at all. I used it to hide "idle" signals for debugging purposes, but it could be worth the attempt to check whether the disk is not draw at all in such a case. Then you could only show green on each of the disks in the stack, which means that whenever an arm is off you see a full green disk, and if all are on, you see nothing in the map. If you then manage to find out how to determine the "bottom" disk in the stack reliably, you could show the red on that one, and than you made it. I.e., you would have a separate script for "bottom disk" signals, i.e., the arm which is placed first, like the main route one, and that one would show the red disk, while the other version shows nothing when at danger. Then, if the user places the main route arm first, and then the others, (and if we are lucky and KRS does not shuffle the stack at random), the red disk will always be at the bottom of the stack, covered by a green one if one of the other arms is off, and visible when all other arms are off.

[Sly401]

I quite like that idea Mark
even if the AI do wait unprototypically at the "on" distant..

The info displayed to a player whilst driving will be correct

Sly

[mbtech22]

I quite like that idea Mark
even if the AI do wait unprototypically at the "on" distant..

The info displayed to a player whilst driving will be correct

Sly

Having carried out several days of testing of my next signal kit with AI traffic, I have decided for now to set the distant signal "Stopping" flag at False for now. This means that the distant will be ignored by AI traffic and they will not be visible in the F5/2D map. The main reason was that on single line working the AI trains were treating the distant as an "intermediate-block" signal and getting onto the single line early. Sorry, Sly.
For those who wish to have the F5/2D indication for distant signals and are not fussed about AI trains the all you need to do is open the relevant .xml/.bin file for the signal and change the "stopping" flag from False to True. Unlike the KRS distants all the entries are contained within the script files to work the map indications.
I have asked RSDL to consider splitting the two aspects controlled by the "Stopping" flag in the blueprint in their next up-grade, but as yet I have not had a response.
The Bracket and Junction Signal kit has been up-loaded and should be available soon.

Mark Brinton

[Sly401]

Hi Mark

No worries thanks for considering it anyway... I don't get much time to drive nowadays anyway

I am sure it will be a big improvement

Thanks again

Sly