Re: [gpsd-dev] [PATCH 1/3] TSIP: improve comments

[Nuno Gonçalves]

I've stopped working with GPSD so I'm abandoning this patch series.
Anyway I'll leave my comments for anyone that wants to pick it in the
future:

I think there is no correct minimum value that we could put, since
I've seen delays from 0.028, but this magic value is just my
experience, for other devices, at present or in the future, it will
change for sure.

For me 0 is the only acceptable value.

Regarding the NMEA stuff, I don't know and I didn't investigate it.

Thanks,
Nuno

On Tue, Jun 21, 2016 at 8:22 PM, Gary E. Miller <address@hidden> wrote:
> Yo Nuno!
>
> On Tue, 21 Jun 2016 19:11:53 +0100
> Nuno Gonçalves <address@hidden> wrote:
>
>> On Tue, Jun 21, 2016 at 6:53 PM, Gary E. Miller <address@hidden>
>> wrote:
>> > I don't have the NMEA doc, but one reference says only 12 PRNs are
>> > allowed in that sentence.  Another says 14.  We'll need to check
>> > that. Also, will there be GLONASS, BeiDou or other sats in there?
>> > Then maybe it should be a GNGSA.
>>
>> Yes, probably all L1 services (GPS, QZSS, SBASs, and Galileo) and
>> GLONASS.
>>
>> I have to look where is the NMEA being created. This will probably
>> take a while.
>
> If we can resolve the max count in the GPGSA I could live with that.
> There seems to be no consistency about usage of the xxGSA's.  So a bigger
> problem than just TSIP.
>
> Here is how Skytraq does it, others are wiildly different:
>
>         "Used to represent the ID’s of satellites which are
>         used for position fix. When both GPS and Beidou
>         satellites are used in position solution, a $GNGSA
>         sentence is used for GPS satellites and another
>         $GNGSA sentence is used for Beidou satellites. When
>         only GPS satellites are used for position fix, a
>         single $GPGSA sentence is output. When only Beidou
>         satellites are used, a single $BDGSA sentence is
>         output. "
>
>> > So what is the delay you see?  Can you identify your model and put
>> > in the fudge you see automagically?  If not can you add in your
>> > observed fudge as a comment?
>>
>> I see delays from 0.028 to 0.5s, depending on device, as specified in
>> the comments, depending on device, number of visible SV, and probably
>> firmware versions.
>>
>> I don't think it is reliable to assume any value, so 0 looks the
>> neutral assumption.
>
> A minimum value would be nice, but if we cant solve it, ttthen documenting
> is a start.
>
> RGDS
> GARY
> ---------------------------------------------------------------------------
> Gary E. Miller Rellim 109 NW Wilmington Ave., Suite E, Bend, OR 97703
>         address@hidden  Tel:+1 541 382 8588