Track "Connecting" Straight Lines

[Grasscatcher]

Here's a post I made over on the Montana wiki:

 

Since I don't get the straight lines because I "clear current track" and save, I went back and checked a bunch of downloaded Archive files.

Voila! straight lines !, so I did some investigating.......

 

The "saved track files" are totally clear. The archive files for that same date start off with a straight line! Those straight lines "go back" to ("at least Very close") the location where you last turned the unit off. They don't actually "connect" to the previous track....at first, I thought they did until very close inspection.

 

Sometimes it's only a single point or few points and sometimes it's a few more, depending on how long it takes to lock onto a good fix. Even with the Archive period set at daily, when you turn your unit on for the first time of the day, it starts logging (it turns out that they are lousy) points based on where it was last "ON".....until it gets a sat lock to update it's position. Same concept as when you move a long distance between GPS usage and it takes longer to lock...... It does the same thing "during" the day but to a lesser degree because it locks on quicker.

 

You can figure it out by zooming in on your computer.

 

The difference is that NOW , with Archiving, those "first" lousy points get "captured" for posterity, and are included in the archive file. It is NOT including random points logged the previous day in today's tracks, as some think. It is just "starting" with it's "last known" location until it gets enough sat data to correctly determine it's actual location.

 

Why is it different now than in the past? Because new units take much longer to initialize at startup, They have to check all the gpx files for changes, load maps, etc etc. All the while, they start logging points based on previous location.....That's the same as it's always been, just a longer process now.And....those points get saved in the archive.

 

So......power cycling all the time actually makes the situation worse. It is constantly having to start from a different location than it's memory says it was/is. The slower the startup time, due to GB after GB of maps, and tons and tons of (mostly worthless) Archive files, and "elebenteen thousand" Caches......the more bogus trackpoints get logged before a good fix is acquired and then an accurate location determined.

 

If you turn your unit off one night and don't turn it on until the next day, those points and straight line are included in the archive file for that "next day" (at the beginning)....because that's when they were logged (at the first "ON")

 

That also explains why I DON'T get the lines and others do....because I ALWAYS "clear current" before logging. and dating all the way back to a 12XL I have ALWAYS turned the unit(s) on for a "significant" period of time to get a solid lock first and, even then, do a "clear" to get rid of the "lousy initializing data" before starting to log a track.

[+Red90]

That is the "other" thing... It should not start recording data until is has a "reasonable" satellite fix. All easily rectified with a couple lines of code...

[Grasscatcher]

I know I have never had, and I doubt if you have either, a GPS that acquired a totally accurate lock immediately on "power up". It just doesn't happen.....

 

With all the gpx checking, and map loading, and other tasks, the starup times are much slower than before.

 

It has to start from somewhere (last known location) or otherwise it would be like a "First time acquisition" every time. Long.......

 

Reseting trip data and clearing Current track are so simple a fix that I can't understand all the hoopla.

[+Red90]

No. It is simple. It waits to start the track log until the fix is reasonable and does not record when it is not. It just goes by the accuracy value.

[tr_s]

No. It is simple. It waits to start the track log until the fix is reasonable and does not record when it is not. It just goes by the accuracy value.

 

Absolutely. Quite a few Android logging apps, GPSLogger is one of them, have exactly this setting: wait until accuracy <x before logging. It works just fine even if the logging interval is set so large that the GPS module turns off between log points.

Edited February 14, 2013 by tr_s

[John E Cache]

I use Google's "My Tracks". It has a button where you can turn on the GPS and get a fix before you start a track.