Hi Matt
I have the same problem than Hutima in this Support Ticket. Do you have a solution now for this problem?
Delete data points - Features Discussion - Big Paua Support
It looked like this (12 months :-).
Now I deleted some entries and tried several things and now Home looks like this. What’s the difference, when the icon before “Home” looks like this (green). Normally, the icon is black, only this entry I modified looks green.
I was outside and no back home it looks again 11 months and 30 days. Do you have an idea, how I can delete only this wrong entry from the database? I tried so many hours yesterday and at the end it looked good, but now still the same problem.
@matt @Hutima : Please help me.
When I click on split segments it looks like this. The changes I made was on 14:20, how can I delete this entry?
I replied in the other thread, but yeah I think the best option is to either edit the backups JSON to flag the items/samples as “deleted” or to correct their times, then reimport those JSON files. Fiddly work, so feel free to ask questions and I’ll assist as I can if you go that route!
For the green/grey icon, each visit in the timeline has its icon coloured by the preceding trip item. So the colour doesn’t actually mean anything about the visit itself, but instead the trip that came before it (green for walking, black for possibly car). It’s just a meaningless design decision, because colouring the visits differently made the timeline view look too busy and confusing.
Hi Matt
Thanks for your answer. I had a look at the JSON Backups. Here I can see the wrong file with the data from 2026. How does it work, to correct it? I have to download the newest JSON File and have to change it? How can I change the file and import again?
Hm. Those ones look like possibly daily auto export files. Instead you’ll want the ones in the Arc App/Backups/LocomotionSample folder, named by week of year.
You’ll want to find the correct week file, decompress it (double tap should do that), then open up the JSON file in an editor and hunt about for the incorrect dates in there and correct them. It’ll be fiddly work! Needle in haystack stuff.
You’ll need to update the “date” text for each sample with wrong date. They’ll look something like this:
“date” : “2025-03-10T00:00:13Z”
You’ll also need to update the “lastSaved” values to be something recent, so that Arc’s importer sees the data as newer than what it has in its database. So for example with this:
“lastSaved” : “2025-03-10T00:00:37Z”
I’d change it to be today’s date.
Then you will need to copy that JSON file into Arc App/Import/LocomotionSample (you’ll probably have to create that LocomotionSample folder in there).
Then when you open the File Importer in Arc’s “Backup, Import & Export” settings view it should show up in the list of samples files. You can then tap on it and it’ll get imported.
Only the samples in the file that have a newer lastSaved will be imported, so the rest will be ignored. Once that’s done your Arc database should have those samples with corrected date. From there… hopefully the processing engine will clean up the rest! Though I’ve never dealt with this situation before, so… definitely let me know how it goes. There might be extra steps I haven’t anticipated.
I don’t know if I should do it, when you are not 100% sure if its work. The datas on arc where 2 or 3 days completely wrong. Now, the last days it looks normal. Could it be, that I only have a few day problems with this wrong date or do you mean it’s possible that I will get more issues through this date change also in the future?
The import will only change samples that have newer lastSaved date than what’s in the database. So it will be isolated to only those samples and nothing else.
I think it would be best to experiment with only a few samples at first. Like 3 in a row, fix their dates (and update their lastSaved dates), then import that samples file.
From there you’d be able to hopefully find those samples inside their timeline item’s Individual Segments view and split them out. Actually it might help to change the activity type of those samples as well, to something that stands out, like “running”. Then they’ll be much easier to find. Once you’ve found them you can tap them to split them out as a separate timeline item, which will cause them to jump to their correct place in the timeline. Then you can change them back to their correct activity type.
A small experiment like that would prove the theory without much risk. I mean, the data is already wrong for those samples, so it can hardly get worse!
