Activity Number in URL vs Upload/Sync Date

mdennie said:

Are the numbers in the URLs (connect.garmin.com/.../4340725770) for activities sequential

Yes, they probably are, though globally for every one of the millions of Garmin users. In other words, you cannot really guess it.

On the other hand the FIT file names in the folder //GARMIN/Activity/ use the date and time in its name, so there it is easier.

mdennie said:

I want to approximate the date of the activity by figuring out the date it was uploaded to Connect.

The approxiation is easy though - open the page https://connect.garmin.com/modern/activities and move the mouse cursor over the names of the activities - you'll see the URLs (including the Activity ID) in the status bar of the browser. So you can easily navigate to the date of the activity you are looking for.

WillNorthYork said:

trux I think the OP wants the date of upload and not the activity date.

OK, sorry, I misunderstood then

Thanks so much! If I understand, this process is capturing the JSON returned by the web service, right?

Interestingly the upload date time matches exactly the activity start time, which of course can't be... I would have to finish the activity before upload.

When the activity is pulled into MapMyRun the date is many years in the future.

I'm guessing the .fit file has a corrupt date, and upon upload Garmin Connect realizes this and replaces the activity start time with the upload time.

Super helpful.

mdennie said:

I'm guessing the .fit file has a corrupt date, and upon upload Garmin Connect realizes this and replaces the activity start time with the upload time.

Check out the node summaryDTO » startTimeLocal in the same document to see whether it is really wrong

... better told, you should actually compare the summaryDTO » startTimeGMT with the metadataDTO » uploadDate, since it seems to be in UTC (at least looking at my own data, I see it there so)

summaryDTO » startTimeGMT and metadataDTO » uploadDate are identical ("2020-01-09T11:28:43.261")

In the splits I see the below. The startTimeGMT matches what MMR shows (2034), and the DoubleValue is a Unix epoch matching that date. Obviously corrupted.

"eventDTOs": [
{
"sectionTypeDTO": {
"id": 43,
"key": "vO2Max",
"sectionTypeKey": "VO2_MAX"
}
},
{
"startTimeGMT": "2034-09-08T19:32:13.0",
"startTimeGMTDoubleValue": 2041356733000,
"sectionTypeDTO": {
"id": 4,
"key": "timerTrigger",
"sectionTypeKey": "TIMER_TRIGGER"
}
}
]

mdennie said:

Thanks so much!

No worries!

mdennie said:

If I understand, this process is capturing the JSON returned by the web service, right?

Yeah.

mdennie said:

Interestingly the upload date time matches exactly the activity start time, which of course can't be... I would have to finish the activity before upload.

Interesting. Checking a handful of valid activities on my end, start date and upload date appear to be correct.

mdennie said:

I'm guessing the .fit file has a corrupt date, and upon upload Garmin Connect realizes this and replaces the activity start time with the upload time.

That's gotta be it.

But at least you have the upload time now, right? It's reasonable to assume that the time/date of the actual activity was pretty close to the upload (same day or within a few days).

Yup! in this case probably the evening before, since the upload time is 6:30 AM and I don't run that early! :-)