Skully

The AI is still flawed, yet it will tell you if track is available or not if you ask for it. To find those flaws we need to ensure proper procedure (https://tutorials.simrail.eu/index-en.html#section30) is used. But as long as players (admittingly) do not use proper procedure then flaws can not be unequivocally be attributed to the AI. Even on a layout as "SG - SP - SK - DGS - DGZ" you can easily create a scenario in which it would appear SK made the foul while in actuality it would have been SG. So in order to find those AI flaws every external influence needs to be contained, hence the suggestion of the system rule.

Looking from the fence towards junction 66 you see the track not lining up correctly.

20230126-buzzer.mp4 I was equally surprised when I heard it the first time. I was actually expecting an external buzzer but apparently these are built into the table. I have no clue where in reality but again I would personally expect it more centerline like below the amp/volt meters. Does anyone have a video of these cubes in action? I would love to see those.

As I read Ir-1, the usage of telephone is mandatory to control the traffic flow (for at least single track operation). The usage of the interlocks is only to ensure safety but not traffic flow. It is not deemed a mere formality hence my suggestion to have the system rule.

According to the Ir-1 instruction manual: So there. 😁 Well according to Google it translates as: § 24 sections 8-9 are the telephone instructions that go with it. See https://docs.google.com/spreadsheets/d/1g-uorn6BElwUeoPNBQ-GZ-z9GR6mE9Y2l46X8VIvZZA/edit SimRail is a consumer version of the SimKol training simulator so every regulation is taken / implemented literally.

I think a lot of stations are modelled as they were in 2016. There are some giveaways in the game which explicitly state this year. 😄 I don't mind leaving the time period a bit undefined as it gives more leeway for different type of running stock (and stations).

According to https://semaforek.kolej.org.pl/wiki/index.php?title=Będzin it should only service line 1 on radio 2. You may want to file a bug report.

You can actually shunt it forward (using substitute) behind signal A and then shunt it back behind signal H. Tried and tested in this build. 😄

I think what he's saying is if we manage to screwup Sosnowiec Południowy then it'll get unlocked. Ehrmm... challenge accepted? 😁

These bits and bobs are the type C interlocking. You can let loose Google translate on https://www.bsk.isdr.pl/srk_blokp.php to get a bit more information.

I put it up in question 2 to ensure I remember. 😁

I can't add any more options to the poll. It only allows for 10 options. "I can only see 9." you say. Correct, I deleted option 6 due to a typo but it apparently didn't restart the count. 🤪 I'll add it next week and, if needed, remove other non-popular options.

Well uetam gave the hints straight away, so Dąbrowa Górnicza Wschodnia has been added to the roster. 😄 Dąbrowa Górnicza Ząbkowice was also briefly seen on the API so this could be construed as a very strong hint to what the devs are working on. 🤫 Anyway, same deal as last week. I'll put up a list and if you wish to see something added toss it up here and I'll add it.

We have had plenty of lockups around Sosnowiec Główny and Sosnowiec Południowy to come to the conclusion that rookie dispatcher keep stuffing trains down a single track irrespective of clearance. I think we need a system rule that prevents setting a route to such a single track without having obtained clearance first from the other dispatcher. Hopefully this will prevent a lockups, at least it'll teach rookies how single line operation should work. The system rule message must be descriptive enough to ensure people understand what they have done wrong and how to correctly proceed.

On EN1, this time arriving at a platform in Sosnowiec Główny. Time stamp is in CET. AFAIK no other players were meddling with doors as I was just arriving after a straight through at Będzin. [2023-01-20 16:03:17]: Disconnecting connId=0 to prevent exploits from an Exception in MessageHandler: ArgumentOutOfRangeException Index was out of range. Must be non-negative and less than the size of the collection. Parameter name: index at PeopleLeaveTrain.SelectPeopleToLeaveTrain (System.Int32 numberPeopleLeavingFromGivenPart, System.Collections.Generic.List`1[T] peopleInGivenPart, SingleNavLink doorToFollow, MovePeopleAtPlatform movePeopleAtPlatform) [0x00000] in <00000000000000000000000000000000>:0 at PeopleLeaveTrain.agentLeaveTrain (MovePeopleAtPlatform movePeopleAtPlatform, System.Collections.Generic.Dictionary`2[TKey,TValue] doors, System.Collections.Generic.Dictionary`2[TKey,TValue] peopleLeaveTrains) [0x00000] in <00000000000000000000000000000000>:0 at MovePeopleAtPlatform.RpcPeopleDisembark (NavMeshLinkController train, System.Collections.Generic.Dictionary`2[TKey,TValue] peopleLeavingTrain) [0x00000] in <00000000000000000000000000000000>:0 at PeopleTrainSync.UserCode_RpcPeopleDisembark__String__List`1 (System.String platformGuid, System.Collections.Generic.List`1[T] peopleCountInfo) [0x00000] in <00000000000000000000000000000000>:0 at Mirror.NetworkIdentity.HandleRemoteCall (System.Int32 componentIndex, System.Int32 functionHash, Mirror.RemoteCalls.RemoteCallType remoteCallType, Mirror.NetworkReader reader, Mirror.NetworkConnectionToClient senderConnection) [0x00000] in <00000000000000000000000000000000>:0 at Mirror.NetworkClient.OnRPCMessage (Mirror.RpcMessage message) [0x00000] in <00000000000000000000000000000000>:0 at System.Action`1[T].Invoke (T obj) [0x00000] in <00000000000000000000000000000000>:0 at Mirror.NetworkClient+<>c__DisplayClass47_0`1[T].<RegisterHandler>g__HandlerWrapped|0 (Mirror.NetworkConnection _, T value) [0x00000] in <00000000000000000000000000000000>:0 at System.Action`2[T1,T2].Invoke (T1 arg1, T2 arg2) [0x00000] in <00000000000000000000000000000000>:0 at Mirror.MessagePacking+<>c__DisplayClass7_0`2[T,C].<WrapHandler>g__Wrapped|0 (C conn, T msg, System.Int32 _) [0x00000] in <00000000000000000000000000000000>:0 at System.Action`3[T1,T2,T3].Invoke (T1 arg1, T2 arg2, T3 arg3) [0x00000] in <00000000000000000000000000000000>:0 at Mirror.MessagePacking+<>c__DisplayClass6_0`2[T,C].<WrapHandler>b__0 (Mirror.NetworkConnection conn, Mirror.NetworkReader reader, System.Int32 channelId) [0x00000] in <00000000000000000000000000000000>:0 at Mirror.NetworkClient.UnpackAndInvoke (Mirror.NetworkReader reader, System.Int32 channelId) [0x00000] in <00000000000000000000000000000000>:0 at Mirror.NetworkClient.OnTransportData (System.ArraySegment`1[T] data, System.Int32 channelId) [0x00000] in <00000000000000000000000000000000>:0 at kcp2k.KcpTransport.<Awake>b__22_2 (System.ArraySegment`1[T] message, kcp2k.KcpChannel channel) [0x00000] in <00000000000000000000000000000000>:0 at kcp2k.KcpClient.<Connect>b__7_1 (System.ArraySegment`1[T] message, kcp2k.KcpChannel channel) [0x00000] in <00000000000000000000000000000000>:0 at kcp2k.KcpConnection.TickIncoming_Authenticated (System.UInt32 time) [0x00000] in <00000000000000000000000000000000>:0 at kcp2k.KcpConnection.TickIncoming () [0x00000] in <00000000000000000000000000000000>:0 at Mirror.NetworkLoop.NetworkEarlyUpdate () [0x00000] in <00000000000000000000000000000000>:0 [2023-01-20 16:03:17]: Client disconnected without intent! Going back to main menu! SimRail_log_2023-01-20_13-09-11.txt

It is already well established that people would like to see private servers to run by their own rules. This would even allow for shunting and coupling within the current engine (as we saw it in the early stages of playtesting). Although the AI is not ready for that (yet). What is also established is the fact that SimRail hasn't been designed with privately hosted servers in mind. Just read through: Furthermore the public servers should be able to run with as little moderation as possible to keep operational cost down. I think once dispatcher XP and levelling comes into play we can propose a strike system. Something already on the radar. Once all of that is dealt with maybe private hosting can become a realistic goal again. Alternatively maybe public expert servers could be put into operation which do allow for more maneuvers.

It can be fiddled with, but what does it all mean? And have the individual settings been implemented in the simulator?

20230119-station.mp4 I knew SimRail had grand ambitions. 😁

Like I said on discord for this occasion and others as well, any train run from one valid stopping position to the next. Irrespective of whether it'll actually stop. A valid stopping position means any other train can pass. Assuming 42936 entered the track Sosnowiec Kazimierz first with 42186 wanting to follow. Either 42186 would have been held at Sosnowiec Dańdówka because it could not reserve a path to Dąbrowa Górnicza Strzemieszyce (Sosnowiec Kazimierz has one track reserved for 42936 which makes the other not a valid stopping position) if 24933 already left to Sosnowiec Kazimierz (and thus holding a reserved path to SK). Or 24933 would not leave Dąbrowa Górnicza Strzemieszyce because 42186 has reserved a path from SD to DGS. This last one would also keep 42936 in place at SK (again because 42186 has reserved SK to DGS). The current AI implementation does not look ahead. So 42186 is allowed to follow 42936 (or vica versa) without much further checks. I suspect it simply stops checking when it sees another train heading in the same direction. If so then this is bad. Anyway at the end of the day overtakes such as 42186 over 42936 should certainly be possible for the AI to do. Otherwise we get interesting delays.

Maybe we could have our steam profile logos graffitied on the train?

Some playtesters know of little bugs that are still in there. 😁 If you drive past red you get teleported back to the signal.

It is not marked as a planned stop so I rumbled on at full speed. I tried to find it on the OpenRailway map but apparently I did not zoom enough. 😁 So it does exist but is badly scheduled.

The 07s have two traction motors. At throttle 28 one is engaged fully.

Give praise to Friend Computer! Happiness is mandatory! And hope for the best... Blocks are guarded by stop-and-go signals so they can completely fill up back-to-back. Toss up a screenshot of the map next time. Maybe there was some service far out which was considered a higher priority.

These are partial captures of the screen not full screen. I'm running 2560x1440 16:9 and graphically everything looks fine. Ah ha... good point. Found it. So the bug becomes "Wrong language dispatcher manual by default". 😄