Latest XT2/Tread software version

[Lanso]

Try reloading it a few times

It'll start to mess up.

[Regain]

@Lanso I did, a few times and even let Tread sync with the XT2 then set routes running in simulator mode. All seemed ok. I’ve been using this system quite some time now and apart from some teething trouble, it seems to work well.

How do you plan your routes?

[proofresistant]

@Lanso I did, a few times and even let Tread sync with the XT2 then set routes running in simulator mode. All seemed ok. I’ve been using this system quite some time now and apart from some teething trouble, it seems to work well.

How do you plan your routes?

If it works for you, then it's fine, and hopefully it'll stay that way.

Then it doesn't matter how you plan such route; it all comes down to whether the Tread app and/or the zumo XT2 don't like the positions of the shaping points you've chosen.

If you would some fun with this issue, try out the demo route from smfollen, You’ll see what the problem is very quickly.

Re: Tread Changes Exisiting Routes - Shaping Point Issues
viewtopic.php?p=22791#p22791
Attachments
ChathamLoopOriginal.gpx
download/file.php?id=3291
You might have to download the US maps for that, but I think it works without them too—I just can’t remember exactly.

[proofresistant]

iOS is on v3.00.09 updated three days ago and still current.

For whatever reason, updates seem to roll out differently in different places or on different systems—who knows?
I’d known for days that v3.00.09 was available—I’d even found it online—but on my device, it wasn’t updated, and I couldn’t install it fresh either.
I myself had v3.00.09 running for just one day, and then v3.00.11 was already on my Android.

[jfheath]

I looked briefly last night. Android Tread App - I thought it was v 3.00.10, but this morning it reports 3.00.11. Maybe I was mistaken. When did the app start to update itself ?

I looked at the shaping point route that I documented here: XT2-P351c

This was a route created with 8 points, all of which were shaping except the start and end which have to be vias.

The route was created as a gpx file and placed in the internal storage/gpx folder and then imported by the XT2

In that test, without tread allowed to synch, the route points were renamed and moved slightly but the route stayed as it was following the A6 road.
With tread synch active, the route jumped onto the nearby motorway. the same test with via points did not do this.

With the latest version of tread this has changed. As soon as tread was allowed to synch - after a brief delay while it did its manipulation of the route, the route stayed on the A6 rather than going to the M6

But

One of the shaping points had been changed to a via point, and on the tread app on the phone, it didn’t know where the shaping points were. The list view showed them, but when one was selected, the circle highlight picked out the nearest via point.

I deleted everything and tried the same test again. This time I got a different result - the route jumped onto the M6. Again the shaping points were not shown on the app on the phone and selecting them from the list highlighted the end via point.

I got the impression that what had happened was that the route had been changed to the type of route that is produced by converting a track to a trip - ie no route points. Or the type of route that is created after pressing ‘skip’. I have no proof of this but know how to find it. The biggest issue is why it produced a different result the second time. Such routes do not aim for the next route points if you have to deviate, they aim for the point closest to your current position. If you miss a turning to the next via point, the via disappears. Such routes are always prone to RUT behaviour.

But that is just gut feeling based on one test.

The good news is that if you don’t use Tread, all is well (except the renaming of route points)

I’ll look at this some more.

[Regain]

@proofresistant you're right, you should be able to import a route from whatever software you’ve used to plan. I’m just curious whether Basecamp is causing issues. I’m prepared to get blasted by its faithful users now. I used it myself and Mapsource before that, real paper maps before that. All I’m saying is that using MRA with shaping points on the roads I want, it seems to work as advertised.
I’ve even planned on Tread which can be tricky but achievable.

@jfheath I haven’t seen anything like that behaviour. I don’t name my shaping points but do my via points and they keep their names on Tread/XT2. I have had RUT occur on routes imported from SD card but now run a copy as you suggested which cures it, even routing round road closures. You can add POIs either on the route or nearby to mark places and they appear faithfully named too. And like you also suggested, more shaping points to prevent the route snapping to an Autoroute or Motorway which cured it. Thank you.

[jfheath]

@proofresistant you're right, you should be able to import a route from whatever software you’ve used to plan. I’m just curious whether Basecamp is causing issues. I’m prepared to get blasted by its faithful users now. I used it myself and Mapsource before that, real paper maps before that. All I’m saying is that using MRA with shaping points on the roads I want, it seems to work as advertised.
I’ve even planned on Tread which can be tricky but achievable.

I don't believe that Basecamp is the issue - a comment that I make with reservation.

A while back I tested out GPX files that were created with Basecamp, MRA, Kurviger the Tread app itself and one other (Calimoto ??). All of them Exported as GPX files. There are some differences in that some do not include the Ghost points, neither does Tread. Some don't have shaping points (eg Kurviger sets everything to be Waypoints set as Vias - or it did when I tested it, it has chanegd recently). For my tests, all had the same problem.

Basecamp produces a very well behaved GPX file - but the ghost points are a particular feature in Garmin satanavs and routes. MRA emulates these in gpc v1.2 but strips the shaping points. Whatever, for my tests I always assumed that at some point the Zumo would recaclulate, so I force that to happen and work with the results.

There was no noticeable difference in the behaviour. Except MRA routes do not change the names of the route points.

@jfheath I haven’t seen anything like that behaviour. I don’t name my shaping points but do my via points and they keep their names on Tread/XT2. I have had RUT occur on routes imported from SD card but now run a copy as you suggested which cures it, even routing round road closures. You can add POIs either on the route or nearby to mark places and they appear faithfully named too. And like you also suggested, more shaping points to prevent the route snapping to an Autoroute or Motorway which cured it. Thank you.

Via Points tend to have their names changed some of the time. Maybe about a quarter of them - from casual observation, and it appears to be random - But I have noticed the same test on the same Zumo with the same GPX file has changed different route point names in different trials.

RUT will happen on any imported route, but you have to know what triggers that behaviour. For certain, if you skip a route point, the entire route is then recalculated. At that point the recalculation is perfectly OK, but the nature of the route has changed. If you deviate from the plotted line, it does not do the same as it would if the route had been built on the XT2 screen, or if it had been built on the Tread App. Instead, it behaves like a track that has been converted to a trip - ie as if it doesn't have any route points. So it no longer heads for the next route point, it heads for the closest point in the remaining route and heads for that. If it has just recalculated the route and has decided that you have to turn round and go back, then the closest point is behind you. Keep ignoring it and the same thing happens. The closest point of its new route is always behind you. That is the RUT scenario. Furthermore, if you rejoin the route ahead, it will than navigate ahead, but if you miss an instruction to (say) turn left to the next route point it is likely to remove that route point (via or shaping, it doesn't matter) from the route. It just disappears from the map.

The point is that it is perfectly possible to have a trouble free route in this scenario, if you do not deviate from the magenta line or if you follow the instructions completely and/or you do not trigger the complete recalcluation of the route. The problem is that I want to be able to rely on my satnav when I am in places that I have never vsitied before. If I have a route that is getting close to our limit for a day's ride, I don't want to be spending half an hour trying to work out where I am in relation to the route that I plotted and the roads that I wanted to ride. I need to be able to recognise when it has done something stupid, and how to fix it quickly. In the pouring rain, at the end of the day, with cold hands and wet gloves. Cos that is when these things happen.

But when I do my tests, I want to know whether or not a particular issue has been fixed. So I use routes that I know will work perfectly if I build the route using the XT2 screen. Or if I make a copy of a route from a GPX file (ie Saved Routes rather than Imported).

I reckon that identical imported routes should behave in the same way. And for imported routes I know how to provoke them to go wrong - in the same circumstances where 'Saved' routes will not go wrong. The issue is - will the route behave like that if Tread is involved - even if I press Skip, even if it has been synched and re-synched. Even if I change the name by adding a full stop.

This incident with one of the shaping points being made into a via, I could not reproduce this morning. Thankfully, yesterday, I got screen shots, so I know that I didn't imagine it ! I am wondering if I did have 3.00.10 on the XT2 - it was updated only a couple of days ago, and now it is on 3.00.11. I'm wondering whether they spotted the issue, and have updated it to where it was before !!

I did notice that behaviour when Closest Entry Point was first introduced. I was investigating what it did at the time, and no sooner had I worked out what it was doing, they brought out another update. So the first one I tested did nothing more than automatically select the nearest Via Point from the list when starting a route. The next update modifed that to be the nearest Via Point, or the nearest shaping point (if it had been created as a Waypoint). The next update changed it again to be the next via point or the next shaping point. Then it headed for the closest ghost point on the route - which is how it remained - with a few modifications due to the behaviour if U turns were not allowed. But why not get it right before releasing the update ? And the second update I observed suggested to me that the developers were not entirely sure about Garmin's definition of Waypoints, Via Points and Shaping points.

[jfheath]

Here are a few screen shots - taken last night.

I THINK that I had 3.00.10 installed - but cannot be sure.

Here is the route as imported by the XT2 - The route has 2 via points - start and end, and 6 shaping points to keep the route on the A6.
The M6 is closeby. There is a junction just after the start point, and another just before the end point.
Another junction with the M6 is just south of point 04 A6 Shap.
The finish point is at the service station on the motorway, but the shaping points are placed to keep the majority of the route on the A6.
Bluetooth is turned off, so the route cannot synch with the Tread App.

Test 01 All ShpPts Before Synch.png (317.55 KiB) Viewed 653 times

After Bluetooth was turned on, the route eventually synched to the Explore / Tread cloud. And then synched back again.

Good news was that the route stayed on the A6 - previous attempts at this had resulted in the route north of point 04 and the shaping points all moving on to the M6 motorway.
Not so good is that one of the shaping points has been altered to a via point.

The image below shows the altered route with the altered shaaping point - now a Via.

Test 01 All ShpPts After Synch.png (346.86 KiB) Viewed 653 times

Note that this morning when I tried to repeat this behaviour, it did not do this. It moved the points from 04 to 07 onto the M6 motorway.
App version 3.00.11


I like to keep my vias to a minimum - they mark my stopping places so that the Trip data readout at the side of the Zumo screen shows me how far and how long before we have a scheduled stop. Like this - 27 miles to the scheduled cafe break. 47 miles to the end.

ExampleTrip Data.png (157.98 KiB) Viewed 653 times

[proofresistant]

@proofresistant you're right, you should be able to import a route from whatever software you’ve used to plan. I’m just curious whether Basecamp is causing issues. I’m prepared to get blasted by its faithful users now. I used it myself and Mapsource before that, real paper maps before that. All I’m saying is that using MRA with shaping points on the roads I want, it seems to work as advertised.
I’ve even planned on Tread which can be tricky but achievable.

@Regain
What I wrote earlier wasn’t meant to suggest that the planning tool used to create the route might matter.
On the contrary, based on all my tests, the source is irrelevant.
No matter what I used for planning—whether BC, the Zumo XT2 (Tread 2) device, Kurvieger, Tyre2Navigate, or even the Tread app itself— If anything, the results are always comparable: re-imported shapings are shifted.
I strongly suspect that if you haven’t observed this phenomenon yet, it’s not because of the great MRA, but rather because the shapings required for your route guidance are simply not susceptible to it.

One more thing:
I’m not aware of anyone anywhere who has figured out what the actual cause of the shifting of the (edit: no sharp turns) Shaping Points is and knows exactly how to avoid it when using the Tread app.
And so I strongly believe that MRA doesn’t know this either.
And if MRA doesn’t know this, then they couldn’t have invented a solution for it either.

[proofresistant]

...
RUT will happen on any imported route, but you have to know what triggers that behaviour. For certain, if you skip a route point, the entire route
...

@jfheath, I'm not entirely sure about the “any” part from "RUT will happen on any imported route" right now more.
I’ve noticed, I don’t know if it’s always been this way, that the “Imported” flag is set to [False] for routes imported from the Tread app.

PNG-Bild (neu) (4).png (73.22 KiB) Viewed 640 times

And with the “imported” flag set to false, there shouldn’t be a RUT problem, right?