Problem uploading a specific route to Zumo XT

[jfheath]

I had a look at @Thornton's you routes and got the same results. The route had 16 route points, made up mainly of Waypoints. All but one were set as Vias.

Curiously part of his route I can see through my window down the valley. But its not the Yorkshire grit that gets into the workings of the motorbike that is causing this issue.

The route is listed as being available to import, it starts to import and then returns to the screen as if complete, but suspiciously quickly.
No route is found in trip planner.

I separated the waypoint definitions. Waypoints worked, route didn't.
I changed all Vias to shaping. No joy.
I changed maps, recalculated. No.
Changed maps back. No.

Time to start doing a binary search - ie split the route in half to find in which half the fault is located.
So one route ended at a mid point, the other half started at the same midpoint. I did this with the trip list in Bascamp.

Yes. The first half worked. So the fault must be in the second half.
No. The second half worked too.

I'm stuck for ideas.

Something will come to me, like keep extending the first half until it fails.
Is it the number 16. A nice round number for a computer. I doubt it.

[Oop North John]

Not downloaded the route etc, but is the route having the double used waypoint problem I had?

viewtopic.php?t=2449&hilit=route

[jfheath]

I did spot that the same point was used twice. Also spotted that the full waypoint definition name etc was declared in the route as well as in the waypoint section.

I thought that the full waypoint was declared at the head of the gpx file (or separately), and when used in the route it was just the same coordinates and the name. But I might be mistaken about that. It just looked odd.

That double route point would certainly explain why splitting the route worked - the repeated point is at the start and end. Not good routing practice anyway. You never put the start at the start, and you never let the start and end get clise to each other. Habit from 660 days - when the route would finish as you set off. But the same thing can happen if you use all shaping points and recently discovered , if RUT behaviour has been triggered by an imported route being recalculated - route points do not have to be visited and are removed if you miss them!

Thanks @Oop North John . It seems likely. I'll check it out and report back.

John

[Thornton]

I did spot that the same point was used twice. Also spotted that the full waypoint definition name etc was declared in the route as well as in the waypoint section.

I thought that the full waypoint was declared at the head of the gpx file (or separately), and when used in the route it was just the same coordinates and the name. But I might be mistaken about that. It just looked odd.

That double route point would certainly explain why splitting the route worked - the repeated point is at the start and end. Not good routing practice anyway. You never put the start at the start, and you never let the start and end get clise to each other. Habit from 660 days - when the route would finish as you set off. But the same thing can happen if you use all shaping points and recently discovered , if RUT behaviour has been triggered by an imported route being recalculated - route points do not have to be visited and are removed if you miss them!

Thanks @Oop North John . It seems likely. I'll check it out and report back.

John

I will make a couple of changes. There is also a point in the middle that is used twice - Hollingworth Lake.

[jfheath]

I didn't see Hollingworth Lake used twice.

But - I've tried a few things

1. Replaced the end point with a different end point
2. Replaced the start point with a different start point
3. Replaced all of the route points that had a phone number in them. (Like the Coop). These must have been obtained from Garmin's database, possbly because you happened to place the mouse over the location of a supermarket or garage.

All of the above failed.

So I continued with my Binary search - except it was quicker to make on change at a time.

I knew a route using all of the points to Laycock worked. So I systematically went through and duplictated the orignal route.
So I got HB Tour 1, HB Tour 2, HB Tour 3 - up to HB Tour 8 - I did not duplicate the waypoints. They remained as they were - so all 8 routes were using the identical points.

Then stripped out the last 1 the last 2 the last 3 etc route points, exported them all and tried to import them all in one go.

Only the original and HB Tour 1 failed. The others all imported perfectly.

And that is possibly becasue the last and next tolast route points are duplicated. The route thet misses off the Coop and Hollingworth lake and ends with Heptonstall works fine.

(I thought Heptonstall might have been the problem, because the road is closed to non residents - but no, it wasn't that..

One last try now you have spotted that Hollingworth lake is also used twice.....

[jfheath]

Yes that works with new waypoints plotted for the last two waypoints that were used twice.

It is not possible to simply rename the waypoints. The Basecamp Database is a fully normalied one with no unnecessary data duplication - so when you create a waypoint, and copy or use it somewhere else - you simply copy a pointer to the same data - you don't create another identical point. So if you change the name in one place, then it changes the only instance of that waypoint. The place where you used the same waypoint - the name will have changed there too. So renaming one of the two 'copies' of a waypoint in one route, will change the name of both of them !

Digression:
THis is important for other reasons. If for example you have a point somehwere away from your house that you call "Start", and you use that point in almost every route that you create. Lets say that there are 20 routes that use the point 'Start'. Later you plot a route, and you are heading in the other direction - so you create the route as normal, but you then decide to move Start to make it easier to pass through that point when you set off.

But there is only one entry in the database called Start. The 20 places that it is used merely point to where the data is stored. So if you move it, then that point will be moved in all 20 of the routes in which it was used. You get a clue about this - because Basecamp does the correct thing. It know which routes use that point, and it immediatley goes away and recalculates all 20 routes.

If you double click a waypoint in the bottom left pane of the BAsecamp screen (where the routes, tracks and waypoints are all listed), you will get all of the details that are stored witht he waypoint. Address etc. Across the top are a number of tabs. Categories can be used to aid searching when the data is stored on the Zumo. Just make up your own categories. But the tab on the right - References - lists every place inthe Basecamp database where there is a pointer that accesses this waypoints data. Change that waypoints data - particularly the position, and all of the routes listed will also be altered to take on board the new information.

The people that designed BAsecamp really knew what they were doing when it came to database design.

Just for the record - I loaded these routes into the 590 and the 595 as well.

Both devices imported the modified route with no duplicate points instantly, and the route was available in Trip Planner.

For the original route with duplicate points, both devices took a long time to import, stalling for a long time between 20 and 30% and progressing very slowly after that.

The 590 succeeded eventually (5 mins) and the route displayed on the preview correctly and the trip planner seemed to be correct. Mileage and time were close enough to Basecamp's.

The 595 took about 5 mins and then aborted at around 92% with no route visible in Trip planner.

Clearly it was recognised by Garmin that there was a problem with importing routes of this nature. The XT seems to spot it straight away - so I guess the programming includes a check to see if the same waypoint has been used twice. The 590 didn't see the problem, but it still took a long time. The 595 aborted after 5 mins.

[Thornton]

I tested it using exactly the same route as the original (problem) route. The only change I made was to substitute one of the Hollingworth Lake waypoints with a new one I created just down the road.

My new route works fine.

So, in this case, it's not the duplicated start/end point that is the problem, it was the single waypoint "Hollingworth Lake" that was used twice.

From memory, I created that waypoint by right clicking on the BC map and used "New waypoint" to create it. I believe I may also have "moved" it a touch to ensure it was on the road, rather than just off.

Hollingworth Lake.jpg (116.05 KiB) Viewed 846 times

Interestingly, I can see (what I assume is) the original "Hollingworh Lake" on the map.

I have a suspicion that this "bug" is related to the fact I moved an original point, without renaming it, rather than the fact I used the same one twice.
I say that because a) I've been using the same start/end points on a lot of routes without problems, and b) I've also used the same via point many times when a route crosses on inbound and outbound legs.

Also interesting to note that my original route loads perfectly well into my BMW Nav VI, which is, after all, also a Garmin Device.

[FrankB]

Sorry for joining only now, after all the hard work has already been done.

To me it looks the same as this one: viewtopic.php?p=17656&hilit=import+route#p17656

Waypoints should be unique in a route.

Edit: Let me rephrase. Waypoints added as Via points or Shaping points to a route should be unique.

[Thornton]

Sorry for joining only now, after all the hard work has already been done.

To me it looks the same as this one: viewtopic.php?p=17656&hilit=import+route#p17656

Waypoints should be unique in a route.

Edit: Let me rephrase. Waypoints added as Via points or Shaping points to a route should be unique.

Thanks, although I'm not sure I really understand. Seems to me that ensuring Waypoints are unique in a route is more of a workaround than a feature.

This is based on the fact I have many, many routes using the same waypoints within them (eg start/stop at same place), and also the fact the original route of this topic works fine on devices other than the XT.

Therefore, my conclusion is there is a bug in the XT which can be circumvented by ensuring all waypoints are unique within a route.

Would you agree?

[danham]

Another guess... but I think FrankB means no waypoint name should be used twice in a route.

-dan