I came to the conclusion that I was over engineering the problem.

The PIC will be getting a POSIX Time stamp from the Server every 24hours or
less and every so often it has to send data back with a POSIX time stamp.
All I need to do is save the POSIX Time Stamp and increment it with
an interrupt from the RTC every second.

That solves the immediate requirement and works fine.

I also found this code https://github.com/adafruit/RTClib courtesy of
adafruit and the Arduino Community.
I will pull it into my ChipKit UNO 32 and see what changes are required to
make PIC32 Library out of it.

I particularly like the BCD Functions, they probably only work on 2 Digit
BCD but they make my long winded functions look silly:

static uint8_t bcd2bin (uint8_t val) { return val - 6 * (val >> 4); }
static uint8_t bin2bcd (uint8_t val) { return val + 6 * (val / 10); }

Cheers
Chris



On 17 February 2012 15:49, Matt Bennett <mattpiclist@hazmat.com> wrote:

> On Fri, February 17, 2012 4:04 am, Chris Roper wrote:
> > Thanks Alan,
> >
> > I had found that after posting and am busy digesting it.
> > I think I see now how they convert to UNIX Time and handle the Leap Yea=
r,
> > now I need to try and work it back words,  add in the BCD conversions a=
nd
> > take out the DST portion (we work only with UST/GMT for time stamps).
> >
> > I also note my error in the Epoch being 1979 I see now it is 1970 did I
> > confuse it with the GPS Time?
> > With all the software based clocks I have written over the years, alway=
s
> > in
> > BCD, this is the first time I have tried to interact with networks. I
> > didn't realise there were so many conflicting time formats out there.
>
> Not quite conflicting- just different definitions- which has resulted in
> the recent discussions to eliminate the leap-second:
>
> http://www.bbc.co.uk/news/science-environment-16597191
>
> Clocks have been made way more accurate than the rotation of the Earth, s=
o
> they need to be periodically adjusted. I don't think we can predict the
> needs for the leap-seconds more than a year or so in advance, so if you
> want to keep accurate time-of-day, using POSIX time requires a look-up
> table, and may not be accurate with data more than a year or two old.
>
> If you're looking for relative time, POSIX or NTP time works great,
> however, if you're looking for time-of-day, GPS time can be more useful.
>
> Matt
>
> Matt Bennett
> Just outside of Austin, TX
> 30.51,-97.91
>
> The views I express are my own, not that of my employer, a large
> multinational corporation that you are familiar with.
> --
> http://www.piclist.com PIC/SX FAQ & list archive
> View/change your membership options at
> http://mailman.mit.edu/mailman/listinfo/piclist
>
--=20
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
.