Re: [PLUG] gpsd and 2019-04-06

["Eric S. Raymond" <esr@thyrsus.com>]

JP Vossen <jp@jpsdomain.org>:
> I've googled a bit and surfed http://www.catb.org/gpsd/ but I haven't yet
> found any clues about gpsd and the in-progress & impending 2019-04-06
> implosion [1].
> 
> I'm assuming that in context "firmware" means actual firmware on the chip
> itself, and thus below gpsd's level.  But I'm also assuming that gpsd might
> be able to hack in a fudge-factor to handle stupid firmware, since, after
> all, that's kinda what it does [2]. :-)
> 
> My personal interest in this is if my rPi NTPSec server using a 2017
> Adafruit breakout v3 chip is going to lose its mind.  Per [1] it sounds like
> it'll probably be OK, but in general this sounds like a gpsd FAQ to me.  I'm
> getting good data from it but it's not clear to me if it's possible to tell
> from just this how it will react:
> # Captured 2019-02-16_16:20_EST
> $GPGGA,212107.000,4011.3362,N,07529.0403,W,2,09,0.98,69.9,M,-33.9,M,0000,0000*6E
> $GPGSA,A,3,08,27,07,09,23,16,26,18,11,,,,1.29,0.98,0.84*0D
> $GPRMC,212107.000,A,4011.3362,N,07529.0403,W,0.05,117.04,160219,,,D*7B
> $GPZDA,212107.000,16,02,2019,,*5E
> 
> ESR or anyone, clues?
> 
> Thanks,
> JP
> 
> [1]
> * https://hardware.slashdot.org/story/19/02/14/0041206/your-gps-devices-may-stop-working-on-april-6-if-you-dont-or-cant-update-firmware
> * https://www.theregister.co.uk/2019/02/12/current_gps_epoch_ends/
> * https://www.gps.gov/cgsic/meetings/2017/powers.pdf
> [2] http://esr.ibiblio.org/?p=801
> [3] https://www.ntpsec.org/white-papers/stratum-1-microserver-howto/

Sorry, it's impossible to tell what will happen in advance, and GPSD can't
fudge for it.

The reason for both problems is the same.  Somewhere in the firmware of your
device there's a pivot week value.  GPS week numbers below the pivot are wrapped
into the next epoch.  But the pivot week varies by device and is never, *ever*
documented.

Usually it's the week number that the device firmware on the chip
froze.  The problem with this is that there is usually no way at all
to make the chip tell you that date, and certaibly no way that is
device-independent enough for GPSD to use.

Yes, this sucks.  Welcome to GPS-land, where suckage is not an option - it
comes built into the core protocols!

I have a bunch of devices that could go toes-up, too.  Not looking forward
to this any more than you are.
-- 
		<a href="http://www.catb.org/~esr/";>Eric S. Raymond</a>

My work is funded by the Internet Civil Engineering Institute: https://icei.org
Please visit their site and donate: the civilization you save might be your own.


___________________________________________________________________________
Philadelphia Linux Users Group         --        http://www.phillylinux.org
Announcements - http://lists.phillylinux.org/mailman/listinfo/plug-announce
General Discussion  --   http://lists.phillylinux.org/mailman/listinfo/plug