Re: [gpsd-dev] [PATCH 1/1] Initial draft; how to estimate time1 offset

[Sanjeev Gupta]

On Wed, Nov 6, 2013 at 10:30 PM, Eric S. Raymond <address@hidden> wrote:

Sanjeev Gupta <address@hidden>:

> I started off using the term "message-decoded time" from the gpsd web page
> (see: http://gpsd.berlios.de/gpsd.html ).  How would I refer to the NMEA or
> similarly reported GPS solution?

This terminological questionhas come up a couple of times recently - I
had to ask Gary not to use the term "NMEA time" in comments because,
quite often, it's actually UBX binary or SiRF binary time.

I issue a ukase: what you are pointing at is to be referred to in GPSD
documentation as "in-band" or "message" timestamps, as opposed to the
out-of-band PPS pulse. Please specifically use the wor "timestamp"
rather than time.

OK, will use "in-band timestamp".

 

More general question: Why are you separating this stuff from the
Time Service HOWTO?

My initial attempt was just to document Andy Walls' scripts.  I even started in contrib/ .  But the document grew past just a few comments on top of the gnuplot script.  Beside, this document is very ntpd-specific, unlike the more general HOWTO.

What if I continue this as a separate doc, and once it has been commented on and revised, it can be rolled in to the main HOWTO?  Especially if someone can add a corresponding section on chronyd?  It might be easier to proofread as a standalone file.  Sometime before 3.10 you can take this call.

30 secs later ...

AAARGGGHHHHH, again.  Luckily, I did a git pull, and saw that you committed my file already.  I was going to send the entire thing again.  This Internet thing is too fast for me.

--
Sanjeev Gupta
+65 98551208     http://www.linkedin.com/in/ghane