[aprssig] !DAO! format spec?

Tapio Sokura oh2kku at iki.fi
Mon Nov 12 11:56:17 EST 2007

On Mon, November 12, 2007 04:46, Tom Russo wrote:
> No.  This keeps getting stated on APRSSIG as if it were fact, but it is
> not
> true.  Xastir does NOT implement !DAO!.  Nobody has ever coded it up.

Lucky me that I didn't state it as a fact. I was just reading up on Bob's
spec to see if I could produce a patch, but then I ran into some
uncertainties with the wording of the spec.
http://web.ew.usna.edu/~bruninga/aprs/datum.txt is what I was reading and
wondered about the base91 part. I couldn't make out whether the idea is to
convert the base91 value directly to two additional decimals or to
properly scale the base91 value to be between the two steps (1/100 of a

An example: suppose the basic latitude in the packet is N 60 degrees and
10.00 minutes. The base91 value for latitude in the packet is 'T', which
is ASCII 84, so the base91 value is 51. So the method of directly applying
this to the latitude results in N 60 deg, 10.0051 minutes. But if you
properly scale this to the whole range, you end up with 51/91 * 100 =
~56.0439 -> final latitude is N 60 deg, 10.0056 minutes.

The scaling method would seem to be the sensible method to use, because
without scaling you always have a step between 0.0090 and 0.0100 minutes.
With scaling you still have these steps, but they are only 1/10th of the
size each and evenly distributed across the whole range of values.


More information about the aprssig mailing list