The same goes for train logistics mods, for example LTN and Cybersyn. Without a way to know what the new names are the only course of action is to purge invalid data, stop affected trains and inform players that they need to deal with those trains manually.

With this interface change it would be ...

It was made for player interaction and then a lua API was added to interact with that.


If that's the reasoning behind it I think temporary named stops (not rail stops) should work with interrupts because those can only be created via API.

Not that it helps my use-case much because I'm still ...

Here you go. That's the savegame I manually assembled for the initial post, no API calls involved.
Observe how the train's interrupt does not trigger before or after the temp stop when you start the train by sending it to A and let it do it's thing.

Looking at the code I don't see anywhere that ...

I'd like to propose the following signature change to LuaEntity::get_signals() :

get_signals(wire_connector_id, extra_wire_connector_id?, if_modified_since?) → array[Signal]?, MapTick


Read all signals from the selected wire connector.

Parameters

wire_connector_id :: <as before>
extra_wire ...

It seems like in your case at least, the current behavior is actually helpful, right? If an interrupt fired after it departed the temp rail stop and before it went to the temp provider stop, then after the interrupt finished it could go to a different provider than you intended.


Thought about ...

That's no good because the train will still travel all the way to the depot.
And interrupts will trigger once the depot is the only record in the schedule and its initial wait condition is satisfied.

Another problem is that Cybersyn can reuse trains right after a delivery so when a train keeps busy ...

Not triggering from a temporary stop is useful, yes - and also not surprising, considering rail stops are no named stations and cannot connect to a circuit network.

But not triggering from a normal stop when heading to a temporary rail stop is surprising.

How does the "Interrupt other interrupts" setting affect this behavior?

It does not. That was just a test to see if the train manager treats temporary rail stops as some kind of interrupt.

Is this a symptom of the logic assuming every temporary stop is an interrupt stop, instead of reading the ...

https://gist.github.com/user-attachments/assets/84c2653d-ed49-4e03-83a5-be903d857190
(the temporary rail stop is at the orange circle)

Starting the schedule at A, the train will visit A -> temp -> B -> (A -> C -> B)*

Further testing shows interrupts are completely ignored if

a rail stop is ...

Temporary rail stops directly in front of the actual target stops in order to achieve dynamic point-to-point deliveries have always been a workaround. With the additions the rail system has received over the years that workaround has accumulated a lot of drawbacks (for further discussion let's use ...

To rephrase Sopel's "solution":
This system of passing around additional settings through tags probably took some effort to implement but it's been useless the whole time because there's a situation where nothing can be stored in them. Leaving the player with a broken blueprint because it lacks ...

Anyway, my post was meant to be a reply to Harag's statement


What I meant with "debug" is the act of realizing that this is the problem. The engine gives no clue that it disrupted communication and both the sender and the receiver won't fail because neither recognizes that something is wrong ...

It is quite hard to debug this situation. There is zero feedback that an event got dropped in transit.

And even checking all entity references before sending the event is no guarantee that all subscribers will receive the event. One of the subscribers might invalidate an entity in the event data ...

LTN has received the necessary changes and this is now available as a separate integration mod.

I'm just trying to think of an interim solution I can suggest to the SE developers. The proper solution sounds like a larger effort.

EDIT: I played with this a little and getting the rolling stock next to the stop is doable. What's not exactly easy is to get only the tiles under it. Even in my ...

So a workaround could be to reorder source_positions to make sure the rolling stock of a train in state WAIT_STATION that is next to a train stop is cloned first? Would clone_brush get upset if source_positions contains duplicates? That would allow to go in two passes. First find the positions of ...

Issue

The train facing up in this layout is behind the train stop after it is cloned two times via LuaSurface::clone_brush resulting in "no path" after it is switched back to automatic mode. This is because each clone slightly moves the train upwards.

https://i.imgur.com/eiJmSwF.jpeg

How to ...

I noticed that most of the <resource>-rock entities have map_grid=true . I think that's the setting that controls if the resource field is drawn with a checkered pattern on the map and which causes those rocks to sometimes be invisible. I'm not sure if using map_grid=false is the way to fix this ...

Error while running event LTN_Tracker::on_configuration_changed
__LTN_Tracker__/control.lua:164: attempt to index field '?' (a nil value)
stack traceback:
__LTN_Tracker__/control.lua:164: in function <__LTN_Tracker__/control.lua:106>

The elseif on that line should be nested under if data.mod ...

Factorio 0.17.25 breaks pyrawores because it now errors out on duplicate ingredients or results in recipies.
It only reports that iron-concentrate produces 2x dirty-water but once that is fixed it reports more.