ubuntu-phone team mailing list archive

[Thomas Voß <thomas.voss@xxxxxxxxxxxxx>]

On Sun, Apr 26, 2015 at 9:33 AM, Ed Kapitein <ed@xxxxxxxxxxxx> wrote:
> On 04/26/15 06:20, Thomas Voß wrote:
>>
>> Hey all,
>>
>> thanks for the great script, I will make sure that we fold the color
>> coding capabilities into the client executable of the location
>> service.
>> As far as I can see, "green" indicates that the snr exceeds a certain
>> threshold (11 in the awk script). While snr is an interesting value,
>> the more interesting information is "used_in_fix", "has_almanac_data"
>> and "has_ephimeris_data". Even if multiple satellites are visible
>> (with a sensible snr), only
>> those for which almanac and ephimeris data is known are considered for
>> calcucating the fix. http://en.wikipedia.org/wiki/Time_to_first_fix
>> provides a good overview.
>>
>> With that, green should probably refer to satellites that have
>> "used_in_fix" set to 1.
>>
>> The theoretical minimum for a first fix after cold start is 12.5
>> minutes, which is rarely the case and ttf usually takes longer in cold
>> start scenarios.
>> We take multiple measures to minimize this time, but it seems like we
>> are experiencing issues in doing so. It would be really helpful if you
>> could provide the output of:
>>
>>      sudo GLOG_v=1000 GLOG_logtostderr=1 ubuntu-location-serviced-cli
>> --bus system --test
>>
>> Feel free to share privately with me as the logs would reveal your
>> position. Source for the test is in [1].
>> Please note that you should restart your device after the test has
>> been completed.
>>
>> Thanks a lot & once again thanks for the scripts,
>>
>>    Thomas
>>
>> [1]
>> http://bazaar.launchpad.net/~phablet-team/location-service/trunk/view/head:/src/location_service/com/ubuntu/location/service/runtime_tests.cpp#L74
>>
>> On Sat, Apr 25, 2015 at 9:34 PM, Ed Kapitein <ed@xxxxxxxxxxxx> wrote:
>>>
>>> Your welcome!
>>>
>>> I placed a new tarball at the same url, that will show the green
>>> satelites
>>> on top and the red ones at the bottom.
>>> Feel free to test it.
>>> I have 9 visible satellites outside, and still no fix ?!?!
>>> I hope the next OTA update will fix this.
>>>
>>> The web page will reload itself every once a minute, feel free to tweak
>>> the
>>> refresh value in the awk file.
>>>
>>> Kind regards,
>>> Ed
>>>
>>>
>>> On 04/25/2015 09:12 PM, Matthias Apitz wrote:
>>>>
>>>> El día Saturday, April 25, 2015 a las 05:40:32PM +0200, Ed Kapitein
>>>> escribió:
>>>>
>>>>> I wrote a little script, that will convert the info of the satellites
>>>>> into a web page,
>>>>> which you can view on the BQ itself.
>>>>> ( landscape works best )
>>>>>
>>>>> -1 download the tarball at http://www.kapitein.org/sats.tgz
>>>>> -2 extract it in the homedir of phablet ( /home/phablet )
>>>>> -3 run the script: /home/phablet/scripts/gps/scripts/sats2html.sh &
>>>>> -4 after a minute, open the browser and point it to
>>>>> file:///dev/shm/sats.html
>>>>>
>>>>> I hope this will help someone.
>>>>
>>>> I have installed and tested it: works fine, at least I have a table of
>>>> red values of satellites (I'm sitting inside, outside it's raining) :-)
>>>>
>>>> I will map the command to start the script on some F-key in the
>>>> terminal-app, but not in background, because it would cause a constant
>>>> power-drain.
>>>>
>>>> Btw: What is the best way in the BQ browser of page reload?
>>>> Thanks
>>>>
>>>>          matthias
>>>>
>>>
>>> --
>>> Mailing list: https://launchpad.net/~ubuntu-phone
>>> Post to     : ubuntu-phone@xxxxxxxxxxxxxxxxxxx
>>> Unsubscribe : https://launchpad.net/~ubuntu-phone
>>> More help   : https://help.launchpad.net/ListHelp
>
> Hi,
>
> Is there a way to read data from the GPS sensor itself?
> ubuntu-location-serviced-cli must be getting its data from somewhere
>
> Yesterday it had no satellites in sight, despite all location services set
> to on.
> And after i left it on overnight, in front of a window, it still had no
> satellites in sight.
> If i encounter the same problem again, i'll send you the logs, for now it is
> working again, after switching off and on all location services.
>
> The values in the awk file are pure guessing, no logic there :-)
>> 11 = good reception  = Green
> < 11 = poor reception   = Blue
> <   1 = No reception     = Red
>
> If you know better values for good and poor, please let me know.
>

I would propose to mark all satellites with
"have_almanac_data"/"have_ephimeris_data" as green, yellow for
satellites without that data with reasonable snr, and red for any
other reported satellite.

Cheers,

  Thomas

>
>