Page 1 of 1

Weird Behaviour near Via Points.

Posted: Fri May 26, 2023 11:14 am
by jfheath
This isn't a complaint or a cry for help. It is just for amusement now I have nothing better to do - recent observations.

I have noticed recently when looking at the RUT behaviour, that sometimes routing behaviour seems very odd, especially close to Via Points.
Here are a few examples that I spotted and captured. I haven't got a clue what is going on here.

1. Near Harewood House Via Point.

49781.png
49781.png (38.1 KiB) Viewed 3019 times


I spotted this after trying to find whether I could Force the XT into a RUT loop, after changing the mImport Byte to 0. I had previousy tried the same route a number of times, and unless the route was created on the XT, A RUT scenario always developed. This route with the nobbled Byte worked perfectly.

I checked it out when I got home. Loaded the same route with the un-nobbled Byte anf foced it to recalculate by skipping the first route point after the start. It displayed the same thing. A loop around a crescent. So it has nothing to do with RUTs or nobbled mImport bytes. There was no demand to turn rigt (I was travelling from the top of this map) to the via point flag. There was no demand to turn right - and the magenta line was a darker shade - as if it wasn;t part of the route at all. I checked that it was still heading for the Via. It was.

Curious.

----------------------

2. Near Hawes. This was during the Yeehaah route when I tested the nobbled Byte for the first time - which was successful.

492347.png
492347.png (76.67 KiB) Viewed 3019 times

The Via Point was marked ahead - my exit via point from Hawes, about 2 miles behind this screen shot. I am travelling from bottom right to mid left, and have yet to reach the Via Point. When I did reach it, it asked me to turn right. A clever trick if I can do it - a steep roadside banking, a dry stone wall and a field full of sheep. There in no road there, and never has been. For some reason it is tring to get me to go a short way down the road off the B6255 - Lanacar Lane, and then turn back to continue on the road that I am on.

As it happened, I could not go any further the road ahead was closed - some poor soul had come off his motorbike and ambulance and police were in attendance. We had to double back into Hawes and go up Lanacar Lane anyway. Did it know ? But the routing across the field is weird.


-------------------------

3. Same Via Point - Previous Test of RUT behaviour.

512529.png
512529.png (71.59 KiB) Viewed 3019 times

This is the same Via Point on a completely different day and a completely different test. I had plotted a BAsecamp route which BC achieved in the way that I wanted to go with just 2 Via Point other than start and finish. My intention, and the way that BC had plotted it was to first ride to Hawes, and then through this Via Point a couple of miles outside Hawes. When I skipped a very early Via Point (before the two points I mentioned), the route recalculated and it wanted to take a faster way to reach 13 XT3 04 B6255 from Hawes. It was going to approach it from the other direction.

I stuck to my planned route and eventually, the XT gave in (or so I thought). It plotted a route in the direction I was heading to the Hawes Via Point. I knew that something was odd with the milage it was reporting. (see next screenshot), but tat this stage wasn't sure what. I pressed skip in order to see where it thought that it was heading. Its heading for the correct Via.

526070.png
526070.png (82.78 KiB) Viewed 3019 times
The screenshot above is AFTER I passed the Via Point. There was no announcement, and the flag stayed on the screen. The Trip Data on the right reveals all. The Via Point that it is heading for (the one just behind me) is 30 miles ahead and it will take me 46 minutes to reach it.

What it is doing is heading for Ingleton 15 miles away, where it will find a crescent to turn around (I had U turns disabled), and come back 15 miles to pick up the via point from the other direction.

I'm not sure what is going on here - it may be aiming for its original route and there aren't any places to turnround without doing a U turn. I really hope it has not marked the Via as needing to be approached from the correct direction. I think it is RUT related - something which we now know how to avoid.

------------------

Re: Weird Behaviour near Via Points.

Posted: Fri May 26, 2023 9:23 pm
by rbentnail
Your post reminds me of some of the routing I encountered on my 595 and XT in our North Carolina mountains. Directions to go across creeks and streams, into pastures, and down railroad tracks where there obviously were no roads. Goat and cattle paths maybe, but no roads. That's when I switched the 595 from motorcycling to Automobile.

I still see things like this occasionally in Base Camp when moving a point from one road to another at an intersection. I sometimes get a loop on-road or a sharp triangle off-road, either always leads me back to the original route.

Re: Weird Behaviour near Via Points.

Posted: Fri May 26, 2023 10:04 pm
by jfheath
Its most odd, but it is of little consequence really. On the same run i came across another Via. Passed it. I don't remember it announcing - it may have done - but it then popped up the skip window with the Message 'Skip?'. No name. Most odd. But again, it didn't affect the route in any way.

Re: Weird Behaviour near Via Points.

Posted: Sat May 27, 2023 6:48 am
by FrankB
Hi John, I dont pretend to have the answer on this. But I sometimes see things that resemble this. I think it's related to moving a point.

To show you what I mean, here's what you hypothetically did on your first screenshot:
- You added the via point initially on 'Spring Gardens'
- Afterwards you zoomed in, and noticed it wasn't on the A61.
- You moved that via point by holding down ALT.

I get around this by adding a new point, and deleting the old.

In the past I looked at the resulting GPX, and noticed that when you move a via/shaping point, Basecamp does not (always) recalculate the Subclass.

Re: Weird Behaviour near Via Points.

Posted: Sat May 27, 2023 10:14 am
by jfheath
That's an interesting thought.

I don't think that I did that though.
The particular route that I chose was created with Waypoints - ie a predefined point that is saved before the route is created. I know you know what a Waypoint is, but anyone else looking in may think that a waypoint is an ordinary route point. For Garmin it is more than that.

I used waypoints for 3 reasons.
  1. I wanted ot be able to duplicate the route exactly on the XT to compare the behaviours between what we now know as being Saved or Imported.
  2. I didn't want the names to change from what I knew - if something doesn't work as planned it is handy to be able to select 'Skip' or 'Edit Route' to see exacly what the XT is aiming for next. (Edit Route shows the next Via and the next Shaping point)
  3. When testing this route again, I wanted to be sure that there was nothing different about the route points that could influence the result.
I didn't know that you could use the alt key - if I want to move a point I right click and chose 'Move Waypoint' or 'Move Route Point'.

The funny tour round Spring Gardens is not there on the route that is first created as shown on the XT screen. It only appears on recalculation after skipping the firts route point about half a mile after the start.
What is odd about it is that the magenta route on the A61 is lined with a darker red on each side. The detour around Spring Gardens isn't.

But that struck a chord.

In my last email conversation with 'F' - the technical support person with whom I am exchaning emails - An odd comment was made.
'I have been able to reproduce the issue - somewhat'. And followed it up with something like - I normally cannot reproduce navigational faults under simulation, but on this occasion I have'.

I replied immediately questioning how this can be reproduced under simulation - the RUT behaviour only happens when you deviate from a route that the XT has previously recalculated. I was curious as to how they were able to change the location of the bike to be off the route when a simulation to follow the route was in progress.

I have also followed up with three emails concerning what we have discovered recently.
  • The fact that the same route built in the XT behaves itself perfectly - in the identical circumstances when a Basecamp Trip fails.
  • The observation that there is a difference in the XT between "saved trips" and "imported trips".
  • We have found a way to fool the XT into thinking an imported trip is saved. (and outlined the method). And that behaves perfectly.
Having seen this little loop and one or two others - all of which are created when the route is recalculated after Skip, I am wondering if what they are seeing is the loop and they still have not observed the RUT behaviour. I was expecting contact by the end of this week, and have not had any. I'm not available for the next couple of weeks, although I will pick up emails. I'll get onto the chap that Stu made contact with again, since he offered the help last time, and the response was pretty quick.

I wish I could have a conversation over the phone. It would identify and clear up a lot of mis-understandings. I might get in touch with the front desk and ask what is possible.

Re: Weird Behaviour near Via Points.

Posted: Sat May 27, 2023 10:42 am
by FrankB
jfheath wrote: Sat May 27, 2023 10:14 am I don't think that I did that though.
I am sure you didn't. Just a thought.
jfheath wrote: Sat May 27, 2023 10:14 am I wish I could have a conversation over the phone. It would identify and clear up a lot of mis-understandings.
Or what is very common in the post corona era: Teams. A Teams meeting I would be willing to join.

For your info. I have posted a link on a Dutch forum. Maybe we get some help from there.
https://forum.gps.nl/viewtopic.php?t=54124