It might be a long standing issue which I only started noticing recently. First, I encountered this on Horseshoe Curve, at the namesake location on the outer tracks, thinking it's an old route and something has changed. Today, upon running GWD242 the last time, I noticed that something is odd with the M8s. Decided to quickly fly above and capture it so that I can report. A bit curious that it's only specific couplers (mid-set, white ends). Edit: Easiest to reproduce by starting the Horseshoe Curve railfan scenario (on the free roam tab). It might be interesting to check this with a much older version, Spikee explained how.
I'll check on that. As you've posted it here, we're talking about v75.8a I guess. Just downloaded GWD242. I have no idea what freeroam scenario you're referring to, there's many on my system, but none called "Railfan"... does an M8 run there?
Yes, it's the current live. Might be worth checking in beta as well. Didn't dive deep but it might exist if you simply pull an AI here. On HSC it is also a four track section and it's more severe on the outer track. It is also visible on Blazin's shot. Of course these shots won't survive pruning unless it's not too bad. Look carefully at the gondolas right in front of the player loco, the F7.
What exactly does this image show?!? Can't see anything wrong. Please give simple to follow repro steps. First thing I noticed on the FL9 is the missing cab occlusion. Not defined empty field... Will add one.
Thanks for looking into it. I simply mean how it looks. It only affects the AI. I don't think you can fix it - you can only check with different versions. Maybe it was always like this, I just never noticed. This is for Steve. Blazin's capture shows it's not my PC - I just noticed the gap after the leading GP7, too. It's one of those "once you see it, cannot unsee" things like the UP SD40 doors. There is one for Horseshoe Curve, I missed it, too, before. From the manual: 3.1.2 Free Roam Scenarios (snip) On the Horseshoe Curve Observer scenario you start on the viewing area of the curve where several Artificial Intelligence (AI) trains will drive past. You can choose to stay and watch them pass or click on one to take control. Background. As far as I observed, AI doesn't follow physics, but simply calculates the location of each car. The situation might have something to do with distances. Imagine a curve with two tracks. The inside track is shorter, so the distance between mileposts is shorter than on the outer track. Make it four tracks with a sharp curve (Horseshoe) and the difference can be measurable. It behaves as if the AI on the outer track somehow tries to match the inside track. At certain intervals, the AI accelerates ahead.
I'll now make a test, a new NYNH scenario only featuring that exact M8 service. Loading up original - first thing I do not like - that consist is placed over two signal blocks (though ATC blocks). Anyway, I'll just replicate it as it is and see what happens. ...and its timings are implausible Just placing a player service and the AI M8 - same result. So now I have a small demo scenario to test on v72. Test on v72: same behaviour
Next step, swapped the consist for the Reppo M2 - same result. And the next one - ACS64+Amfleet - also same result! Is it really the curve that's the issue? It seems like you said - as if the distances are greater on the outward curve (then that's an issue with the route builder? The track should be treated independently from the others - it seems the metric is wrong or something like that...) Very interesting. Leaves the question - route building issue or core issue. Both this and HSC are very old RSC routes. I don't have enough route building knowledge to say if there's different ways to make a curve. There's lots of things still left to test for to pinpoint the issue. The gaps happen when the train hits a GoVia instruction with a 20mph speed limit. What happens if we let it travel at full speed? EDIT: Checked, same behaviour. Dispatching is not the issue. All I can say is it always behaved this way but I have never noticed it. (As I'm mostly in the cab)
Thanks for checking. Then it's really just something I didn't notice The distance theory could be verified by placing trains on each track with practically identical instructions. Going to accept then that it is something to avoid capturing. As a scenario author it is best then to run the player on the outer track in similar situations, as I didn't notice this with the player train.
