GPS Date Bug with Samsung Devices

[Alex[_5_]]

I made a flight on Saturday, Jan 28 in Calif. using a Samsung Galaxy
phone with XCSoar 6.2.5.
The file would not process on the OLC. After examining the IGC file,
it was recording the date in the file one day ahead of the actual
date, i.e. it was recording Jan. 29 (and Jan 30 after 0000 UTC). My
first thought was it might be something with XCSoar, but more research
revealed it is actually a bug in the firmware on Samsung Android
devices using the Broadcom GPS module. Possibly related to the fact
that it is a leap year. The bug started showing up on Jan 1, 2012.
It affects any other GPS apps on the device.

So if you are using one of these devices, watch out, your OLC files
may not process properly. There is some discussion of it here:

http://tinyurl.com/7em2qfm

[Dan Marotta]

May or may not be related since I've had no problems with uploading to OLC.

With XCSoar 6.2.4 I've found that my GMT offset (-7) resets to zero every
time the system is restarted. Maybe that only happens with a power cycle on
the Samsung. Is it possible your GMT offset is not correctly set in your
software? I think if my GMT offset is set at zero, the date will be
"tomorrow" since it *is* tomorrow in Greenwich.

Also, I download my IGC file from my CAI 302 (serial to Bluetooth) so I
haven't tried to upload a log generated by the Samsung. BTW, mine is a
player, not a phone.



"Alex" > wrote in message
...
>I made a flight on Saturday, Jan 28 in Calif. using a Samsung Galaxy
> phone with XCSoar 6.2.5.
> The file would not process on the OLC. After examining the IGC file,
> it was recording the date in the file one day ahead of the actual
> date, i.e. it was recording Jan. 29 (and Jan 30 after 0000 UTC). My
> first thought was it might be something with XCSoar, but more research
> revealed it is actually a bug in the firmware on Samsung Android
> devices using the Broadcom GPS module. Possibly related to the fact
> that it is a leap year. The bug started showing up on Jan 1, 2012.
> It affects any other GPS apps on the device.
>
> So if you are using one of these devices, watch out, your OLC files
> may not process properly. There is some discussion of it here:
>
> http://tinyurl.com/7em2qfm

[Alex[_5_]]

It doesn't seem to have anything to do with XCSoar or the UTC offset.
XCSoar
is fine. It's something about the specific device and probably more
specifically the Broadcom GPS module.
Mine is a Samsung SGH-1897 phone with Firmware 2.2 Froyo.UCKB1 In the
discussion I
mentioned, they are speculating something in the Broadcom GPS module
firmware is out
of sync due to the leap year, so it is feeding bogus information to
the NMEA data
stream. It affects any other apps that use the GPS besides XCSoar. It
may only be certain specific models of the Samsung devices and
certain specific GPS receivers that may be in them. But it would be
worth it
to examine your IGC files if you are using a Samsung device with an
onboard Broadcom GPS
module before you make a long flight and then can't use the IGC file.


On Jan 30, 5:38*pm, "Dan Marotta" > wrote:
> May or may not be related since I've had no problems with uploading to OLC.
>
> With XCSoar 6.2.4 I've found that my GMT offset (-7) resets to zero every
> time the system is restarted. *Maybe that only happens with a power cycle on
> the Samsung. *Is it possible your GMT offset is not correctly set in your
> software? *I think if my GMT offset is set at zero, the date will be
> "tomorrow" since it *is* tomorrow in Greenwich.
>
> Also, I download my IGC file from my CAI 302 (serial to Bluetooth) so I
> haven't tried to upload a log generated by the Samsung. *BTW, mine is a
> player, not a phone.
>
> "Alex" > wrote in message
>
> ...
>
> >I made a flight on Saturday, Jan 28 in Calif. using a Samsung Galaxy
> > phone with XCSoar 6.2.5.
> > The file would not process on the OLC. After examining the IGC file,
> > it was recording the date in the file one day ahead of the actual
> > date, i.e. it was recording Jan. 29 (and Jan 30 after 0000 UTC). My
> > first thought was it might be something with XCSoar, but more research
> > revealed it is actually a bug in the firmware on Samsung Android
> > devices using the Broadcom GPS module. Possibly related to the fact
> > that it is a leap year. The bug started showing up on Jan 1, 2012.
> > It affects any other GPS apps on the device.
>
> > So if you are using one of these devices, watch out, your OLC files
> > may not process properly. There is some discussion of it here:
>
> >http://tinyurl.com/7em2qfm

[Max Kellermann]

Dan Marotta > wrote:
> With XCSoar 6.2.4 I've found that my GMT offset (-7) resets to zero every
> time the system is restarted.

This bug has been fixed in 6.2.5:

http://www.xcsoar.org/trac/ticket/1682

If you see bugs, please report them to our bug tracker. Talking about
XCSoar bugs on RAS is not the best way to get them fixed ;-)

Max

[Dan Marotta]

Guess I'm OK, then, since I've disabled the internal GPS on the Samsung and
accept GPS data from the CAI 302 via Bluetooth.


"Alex" > wrote in message
...
It doesn't seem to have anything to do with XCSoar or the UTC offset.
XCSoar
is fine. It's something about the specific device and probably more
specifically the Broadcom GPS module.
Mine is a Samsung SGH-1897 phone with Firmware 2.2 Froyo.UCKB1 In the
discussion I
mentioned, they are speculating something in the Broadcom GPS module
firmware is out
of sync due to the leap year, so it is feeding bogus information to
the NMEA data
stream. It affects any other apps that use the GPS besides XCSoar. It
may only be certain specific models of the Samsung devices and
certain specific GPS receivers that may be in them. But it would be
worth it
to examine your IGC files if you are using a Samsung device with an
onboard Broadcom GPS
module before you make a long flight and then can't use the IGC file.


On Jan 30, 5:38 pm, "Dan Marotta" > wrote:
> May or may not be related since I've had no problems with uploading to
> OLC.
>
> With XCSoar 6.2.4 I've found that my GMT offset (-7) resets to zero every
> time the system is restarted. Maybe that only happens with a power cycle
> on
> the Samsung. Is it possible your GMT offset is not correctly set in your
> software? I think if my GMT offset is set at zero, the date will be
> "tomorrow" since it *is* tomorrow in Greenwich.
>
> Also, I download my IGC file from my CAI 302 (serial to Bluetooth) so I
> haven't tried to upload a log generated by the Samsung. BTW, mine is a
> player, not a phone.
>
> "Alex" > wrote in message
>
> ...
>
> >I made a flight on Saturday, Jan 28 in Calif. using a Samsung Galaxy
> > phone with XCSoar 6.2.5.
> > The file would not process on the OLC. After examining the IGC file,
> > it was recording the date in the file one day ahead of the actual
> > date, i.e. it was recording Jan. 29 (and Jan 30 after 0000 UTC). My
> > first thought was it might be something with XCSoar, but more research
> > revealed it is actually a bug in the firmware on Samsung Android
> > devices using the Broadcom GPS module. Possibly related to the fact
> > that it is a leap year. The bug started showing up on Jan 1, 2012.
> > It affects any other GPS apps on the device.
>
> > So if you are using one of these devices, watch out, your OLC files
> > may not process properly. There is some discussion of it here:
>
> >http://tinyurl.com/7em2qfm

[Dan Marotta]

Sorry, Max. After a career as a systems engineer in the defense industry,
I'm not used to people doing such great work for free! I'll try to remember
and, if I see a bug, write a ticket. Please, keep up the great work. I
love the product!




"Max Kellermann" > wrote in message
...
> Dan Marotta > wrote:
>> With XCSoar 6.2.4 I've found that my GMT offset (-7) resets to zero every
>> time the system is restarted.
>
> This bug has been fixed in 6.2.5:
>
> http://www.xcsoar.org/trac/ticket/1682
>
> If you see bugs, please report them to our bug tracker. Talking about
> XCSoar bugs on RAS is not the best way to get them fixed ;-)
>
> Max