Zumo User Forums

That will be interesting.

I did two more test today. I'll post te results on the other thread.

jfheath wrote: ↑Wed Nov 02, 2022 6:11 am I compared the gpx files for a Basecamp route with the same route calculated by the XT. BC uses far more decimal places for the via points than the XT uses. Ive yet to work out what the difference is in distance, but that may be why it didn't think I had passed through the start point - hence the 'Skip'.

Basecamp shows me five decimal places for coordinates, the XT displays six. The gpx files I have specify coordinates to fifteen decimal places whether created in BC or on the XT.

There are several web sites that will calculate distance between two sets of coordinates. Using https://gps-coordinates.org/distance-be ... inates.php, I picked a random spot (40.048057436943054,-81.979034319519997), entered the "15 place" coordinates, then dropped the last nine places. The difference was too small and the distance was reported as zero. When I dropped the last ten places the difference was 0.91 meters, and dropping the last eleven causes a difference of 7.02 meters. The XT usually reports an location accuracy of about 3 meters.

I don't know for sure how many places the XT actually USES for its internal calculations, but I doubt Garmin would record decimals to 15 digits if it only uses five or six. Even if it uses only six decimal places there must be another reason for the "Skip" message.

Oop North John wrote: ↑Tue Nov 01, 2022 3:14 pm

jfheath wrote: ↑Tue Nov 01, 2022 1:48 pm Skip used to be displayed when a route point was missed completely. I haven't seen that for a long time. I thought that it was a feature that had been removed.

I think I saw "skip" when the XT was getting it's knickers in a twist over going off a Basecamp planned route which had waypoints, and I went another way. But, it wasn't wanting to skip the next point, but the one after.

Looking at one of my routes from when I was in Croatia earlier this year, peak stupid routing problems time, and noticed that the track is broken up, and that it follows the logic of Active Log 2022:06:07 12:59 usually when I was following the instructions, but with an added 01, 02, 03 often when I wasn't and this is where the track is broken. Not always following that logic of the added 01, but it might help spotlighting the areas to look at.

Hope this makes sense!

Peobody wrote: ↑Wed Nov 02, 2022 1:40 pm I will be heading out shortly to do more rides of my test route. I think I will do three, restarting the XT between each one.
1. Use start point as entry point, go off route just before the interim destination point, allow u-turn prompts while logging.
2. Use closest entry as entry point, go off route just before the interim destination point, allow u-turn prompts while logging.
3. Change interim destination point to a shaping point, use start point as entry point, go off route just before the interim destination point, allow u-turn prompts while logging, be attentive to any difference in u-turn prompt behavior.

I do not have anything remarkable to report. The u-turn behavior was the same during all three tests. I never automatically prompted to skip the interim point that I diverted around. The only behavior that I felt should have been different was during test 3. Diverting around the shaping point meant that at some point I was closer to the next point (end point) than to the shaping point yet each recalc resulted in the XT routing me to the shaping point, even when I was twice as far away from the shaping point as the end point. Worthy of note though is that after many u-turn prompts, a recalc selected different roads back to the shaping point. Still, it was a shaping point. The XT should recognize it as such and recalc to the next point once it becomes closer.

As for the Active Logs, I do not see any difference in the logging while on route vs the logging while being prompted to make u-turns.

Oop North John wrote: ↑Wed Nov 02, 2022 4:46 pm Looking at one of my routes from when I was in Croatia earlier this year, peak stupid routing problems time, and noticed that the track is broken up, and that it follows the logic of Active Log 2022:06:07 12:59 usually when I was following the instructions, but with an added 01, 02, 03 often when I wasn't and this is where the track is broken. Not always following that logic of the added 01, but it might help spotlighting the areas to look at.

Hope this makes sense!

Yes - that makes perfect sense. Mine was making an entry recording the date and time. For the broken track, if it had to start a new section with the same date and time, to the nearets minute, id did exactly what yours did - add on extra digits to distinguish them.
I like to keep my logs and I discovered that if I selected all of the fragments in Basecamp, clicked the right hand button (PC user) and chose the option to join the tracks, it listed them in the correct order, asked for confirmation and joined them all into one track - filling in the gaps with calculated speed. There was the option to replace the originals or just save the new one.

Yours is doing exactly like mine was doing. Today on a test ride in the car, I couldn't get it to go wrong in a lcoation where I have seen it go wrong before. Hence my question about the updates from Express.

On Friday I am going to repeat a route that I know has generated this problem. Do exactly the same, see if the results are the same.
Then try a few different things. I'll have to shift the start point closer to the junction and repeate the run a few times.

I feel that I am making some progress with this. A change in behaviour the last couple of days. The real test will be on Friday.

If it works OK, then all is good. If not, I have one of the technical staff at Garmin waiting with an open case for the results of the experiments.

Peobody wrote: ↑Wed Nov 02, 2022 5:29 pm I do not have anything remarkable to report.

When I get a day like that it feels like a major success !

Peobody wrote: ↑Wed Nov 02, 2022 5:29 pm As for the Active Logs, I do not see any difference in the logging while on route vs the logging while being prompted to make u-turns.

That is good. So far that is what I am seeing - which is different from what had been happening before. Something has changed. Just got to figure out what it was. But I'll know more after Friday.

Peobody wrote: ↑Wed Nov 02, 2022 5:29 pm The u-turn behavior was the same during all three tests. I never automatically prompted to skip the interim point that I diverted around. The only behavior that I felt should have been different was during test 3. Diverting around the shaping point meant that at some point I was closer to the next point (end point) than to the shaping point yet each recalc resulted in the XT routing me to the shaping point, even when I was twice as far away from the shaping point as the end point. Worthy of note though is that after many u-turn prompts, a recalc selected different roads back to the shaping point. Still, it was a shaping point. The XT should recognize it as such and recalc to the next point once it becomes closer.

Now - I was part way through writing an reply, and ran out of time. We had a trip into Ilkley planned, so I saved it as a draft. What I was about to say seems to have been proved wrong. I've written the detail of this in here viewtopic.php?p=13057#p13057
My testing report started 1 Nov 2022.

But for completeness. What I was going to say is this.

A 9 point route consisting of Vias and Shaping points, going from left to right : V1 S2 S3 S4 V5 S6 V7 S8 V9

What I believed before today:

If you start the route, Visit S2, the next point on the Zumo list is S3 and it will continue to try to get you there.

If you miss out S3 and rejoin before S4, initially the Zumo will route you back left to S3 - unless you turn right and rejoin the plotted route. Because you are on the plotted route it continues navigating ahead towards S4.

If you aim to rejoin the route between V5 and S6, then the same does not happen. You have rejoined the route in a different section, ie after a Via Point. The satnav will continue to try to get you to go back to S3.

This behaviour is what I understand from your description, and it does not surprise me.

What I saw today suggests something completely different is happening.

I feel pretty sure that this behaviour is new. Two things happened.

I had two routes

Route 1) V1 V2 V3

Went through V1. Deviated to miss out V2 rejoined the route between V2 and V3. The setnav let me continue ahead, and V2 was removed fromt he list of points to skip. These are Via Points . I thought that you had to visit Via Points even if you rejoined the route after the Via Point.
It seems as though this may not be the case.

Route 2. V1 S2 V3

I went through V1. Ignored the turn to S2. The XT plotted a route ahead and although there was an opportunity to do so, it did not try to plot a route to visit S2. Instead it rejoined the route between S2 and V3 and navigated me straight to V3.
S2 was removed from the list of points to skip.

Now this is defintiely different. I did notice 3 weeks ago that a route I was followng allowed me to rejoin the route without having visited the missed points. I had missed out a couple of Vias and a couple of Shaping points. It never asked me to go back, but I stopped near the end of the route and pressed Skip. It presented me with the first point that I had missed. Skip again, it presented me with the next point - even though I had visited it. In fact every single point in the route was still in the Skip list. Nevertheless it was navigating me forward along the route as if I had visited them.

But how this apparent change of behaviour has come about and when .... I can only speculate, which I had better keep to myself, 'cos I have no way of knowing for sure. PM if you want my thoughts on this.

Tried my route to work again which I know the XT does not like if I don't follow it. It's about 7 miles and takes 15 minutes or so. The route includes a shaping point.

Initially I chose to start at the closest entry point, but it failed to calculate, so I cancelled it and chose the end point. Very shortly after it started navigating the route I chose to go right instead of left. And until about 1.5 miles from the end, it tried to go back to the original route, even after I'd passed through the shaping point. After work I chose a go home and set off on basically the reverse of the outbound route.

I emptied the track history and am now using it without a card, and have no other maps than it came with.

Notes:

One track on the way home which has 234 entries, 14 on the way out which has 212 entries after I joined them all together.
Shortly after I ignore the turn / u-turn instruction the sample rate of the track reduces noticeably I think till it gives the next "u-turn in xx". Sometimes there's a definite gap in the tracks.

If I was to hazard a guess, I'd say the XT is overloaded when you go off the route it wants you to and can't handle calculating and recording track data.

jfheath wrote: ↑Thu Nov 03, 2022 3:13 pm Route 2. V1 S2 V3

I went through V1. Ignored the turn to S2. The XT plotted a route ahead and although there was an opportunity to do so, it did not try to plot a route to visit S2. Instead it rejoined the route between S2 and V3 and navigated me straight to V3.
S2 was removed from the list of points to skip.

This was one of my scenarios. One difference was that S2 was up ahead on the road that I turned off. All recalcs routed me back to S2 until I used the skip button to skip it. This was a route created on the XT.

jfheath wrote: ↑Thu Nov 03, 2022 3:13 pm In fact every single point in the route was still in the Skip list. Nevertheless it was navigating me forward along the route as if I had visited them.

I have seen this, sort of. The bike icon moves along the route line on the map as normal but the turn instruction at the top becomes frozen with an instruction related to getting to the first missed point. Verbal navigation is off so I don't know what is going on there.

Oop North John wrote: ↑Thu Nov 03, 2022 5:37 pm One track on the way home which has 234 entries, 14 on the way out which has 212 entries after I joined them all together.
Shortly after I ignore the turn / u-turn instruction the sample rate of the track reduces noticeably I think till it gives the next "u-turn in xx". Sometimes there's a definite gap in the tracks.

My test route was 7 miles one way. I rode it three times which means I road three circles because the start point was 7 miles away from my house.
An active log file for each leg was produced. There is no segmentation. I am surprised by how uneven the track point spacing is. Leg time varies from 0:00:01 to 0:00:10 with one standing out at 0:00:37. The longer times are not associated with u-turn prompts. I am pretty certain that the 37 second standout is associated with being stopped at a traffic signal. Other than that, I do not see any correlation between speed, distance and leg times.

How is it that we can do similar tests and have such different results? (rhetorical question)
Maybe it is unreasonable to compare an XT produced for the U.S. market with one for the U.K.

Peobody wrote: ↑Thu Nov 03, 2022 6:29 pm How is it that we can do similar tests and have such different results? (rhetorical question)
Maybe it is unreasonable to compare an XT produced for the U.S. market with one for the U.K.

I wonder if it's down to the speed settings for the roads being more aligned with your roads?