EIT Path Question and Observation

maruffijd

Active member
First the observation.

I noticed during a recent test of a tower's path operability, that the order the trains request does not dictate the order of path assignment. I have a West end yard area where 2 EB trains are to take a run through a long, (10.7 miles) single line tunnel. Path EB 1 has an ATK take control and proceed as planned. Path EB2 (CR 1) then requests the EB run and gets the obvious wait for the EB 1 to clear. A third train waiting in the East end yard, (CR 2) requests the WB path shortly after CR 1's request. The actual order the trains ran was ATK, then the CR 2, and the CR 1 last despite CR 1 asking before CR 2. Is this normal?

(ATK = Amtraak, CR = Conrail)

Now, my question.

Is there a way to run EB 2 in a way to get it to chase EB 1 while EB 1 has the control of the path? Both ATK and CR 1 are starting from different legs of the yard. The idea I have is I want CR 1 to chase ATK separated by only typical 1 block spacing rules. CR 2 is generally irrelevant for this case.

Tower properties ("?" button) options:

display disk disabled
try requery blocked
Dynamic ownership
trc off
no debug

Path options(all paths same):

clear on drive
no tcb processing
restore JCN direction on travel
no exclusive sets

tower influenced assets on the paths:

Enhanced TRC3 invisible interloc <kuid2:61392:8101:52> controlling the tunnel paths
JR SL(04) 1-H dwarf tall <kuid2:56034:501014:1> entrance/exit signals
JR SL(06) 2-H center wide 20ft <kuid2:56034:501834:2> Turnout protection signals
JR SL(06) 2-H dwarf <kuid2:56034:501028:2> distant signals for above 06 sigs(inside tunnel)
JR SL(04) 1-H dwarf <kuid2:56034:501027:1> RH block signals along tunnel
JR SL(04) 1-H-L dwarf <kuid2:56034:501030:1> LH block signals along tunnel
US&S Point Machine With Lights 1 <kuid2:334896:262122:2> Junction switches

In IT Enhanced Manager rule configuration:

tower paths tab:

type: enhanced
assignment: ai then external
disc: disabled
path requery: disabled
dynamic option: dynamic
trc compatibility: off
debug: no

Path details tab(all paths same):

clear on drive
auto assign only players (will use mission code for player train, AI gets the "assign path" driver commands as applicable)
no tcb processing
reset JCNs
no exclusive sets

ADDED PICS OF YARDS
https://www.trainzportal.com/mytrainz/view_media_post?media_post_id=123615
https://www.trainzportal.com/mytrainz/view_media_post?media_post_id=123616
 
Last edited:
Hi.

Very shortly : currently either standard ITs and EITs (that shares the same internal data structure than ITs) does not guarantee anything about the order of paths activation when several requests are queued in waiting for activation state. So yes it is possible that paths request EB1, EB2, CR1 received by the tower in this order may be granted in the order EB1, CR1, EB2 ...

If you want a more lengthy and detailed explanation, this is due that path requests are stored in the tower internal data by reusing data blocks identifying trains and paths, and that an implicit priority is given by construction to the train having requested a previous path in the tower sometimes ago. The order of paths activation when more than two paths are in conflict depends on the history of requests of the owning trains requesters ...

This is clearly currently a poor implementation as it is not possible to predict the order for paths activations, and next EIT v63 that should be released end of August 2019/beginning of September will implement a new method for storing waiting path requests in a tower that should respect the order of requests when several paths requests are queued waiting for activation. So need to be a little patient, but the problem should be solved in a near future.
Take care also that this will work only inside a tower : in case of several conflicting paths in distinct towers in dynamic ownership mode, the order of path activation will remain random as it depends on when the distinct towers waiting paths thread will be dispatched, which is something unpredictable ...
Take care also that a waiting queue path with all its ressource available will always be activated before any conflicting path which has not all its ressources available.

Hope this helps.
Regards.

Pierre.
 
It helps a lot. I do look forward to the V63's when they drop and don't mind waiting at all. You do the voodoo that you do so well.

I actually noticed the follow on paths released in the odd order due to the third request was indeed available before the second request because of how reset on drive works. (Wiki's are an amazing read, aren't they.)

After doing some trial and error experiments, I have figured out a working method to get the 2nd EB to chase the 1st EB on a 1 block follow on. I split the EB path into 2 chunks. This will release EB 1 at the first signal in the tunnel at 1 block and give the path control to EB2. I just repeat the process for the exit end and both EB trains run the tunnel based on signal spacing rules. There will be no 2nd WB so the dead space going East is not gonna cause a problem.

To prevent confusion, I decided that through the use of strategic "wait for trigger" releases for the trains here, I am going to have an Amtraak going WB first, then when clear, an Amtraak going EB(1) and the player controlled Conrail going EB(2) in that order. This has been successfully tested.

With that said, I will use one of my favorite quotes: "Cave Johnson. We're done here."
 
Back
Top