There is a proposal in to the Plenary to drop WGS84 Ellipsoid and
go back to the FAI Sphere

February 19-23, 2014

http://www.fai.org/component/phocadownload/category/?download=7583:annex-24a-proposal-software-working-group

1. GPS distance measurements

1. 1 Proposal

1. Change Section 7 in its entirety, including all annexes and referenced rule
documents, to the effect that the only earth model used for distance
measurements within CIVL is the FAI sphere.

2. Include the following text in “CIVL GAP”, the scoring documentation for
centralized cross country competitions: “As of January 1st, 2015, all distance
measurements and airspace validations are based on the FAI sphere. From then on,
it is each pilot’s responsibility to use an instrument that indicates distances
to relevant features (such as turn points or airspace boundaries) based on the
FAI sphere, or to adjust their flying to compensate for the differences in
distance calculations between different earth models.”

3. In “CIVL GAP”, reduce the tolerance for turnpoint cylinders to 0.01% of the
cylinder radius, with a minimum of 1m.

Joerg Ewald <<joerg>>
writes:

I sure hope they vote for the proposal, since that will mean that
what you want ("I just want everyone, all software and all instruments to use by
default the same thing.") will finally happen. And not only in Section 7D, but
also 7A and 7B, everywhere in our world.

To put it all in a bit of historical context: Up to 2012, we had a pretty
schizophrenic situation in CIVL: Distance calculations in comps were based on
the FAI sphere, with a huge margin of error to accommodate for people flying
with instruments that use the ellipsoid. For records and batches the WGS84
ellipsoid was used. How this came about, I don't know, must have happened before
I started paying attention to what CIVL was up to.

In 2012 CASI, the group responsible for the General Section of the FAI Sporting
Code (a group mainly consisting of model airplane pilots and parachutists!)
Noticed that different air sports used different methods for distance
calculations. In the true centralized spirit of FAI, this was deemed
inappropriate, and CASI added some text to the General section, stating that
from 2013 onwards, only the WGS84 Ellipsoid must be used.

Soon afterwards, some people saw the "WGS84 ellipsoid is mandatory" text in the
General Section and acted accordingly. One of them would be SeeYou, it seems.

I started working on implementing this in FS in 2013, and soon ran into
problems. There is, for instance, no simple mathematical way to calculate the
distance between a point and a line defined by its two end points on the
ellipsoid. So airspace checks are nearly impossible to do. Over time people
knowledgeable in scoring and software from several air sports came together,
discussed the matter, talked with CASI, an official task force was formed
(consisting of those people knowledgeable in scoring and software) to work out a
technically sound recommendation to CASI. I was a member of that task force. The
outcome of it, in a few words, was "Let the different air sports decide, they
know best what their requirements are, and how to deal with the specifics of
their sport." CASI gladly followed our recommendation and changed the General
Section accordingly. Most involved air sports decided to stick with / go back to
the FAI sphere, since the accuracy it provides is sufficient for the distances
covered by those air sports.

So now we're back in charge. My proposal aims at ensuring that for all distance
calculations in our sport, the earth model best suited for our sport is used.
And that would be the FAI sphere, in my opinion.

Just to be clear. Garmin GPSes use the WGS84 ellipsoid earth model
and so in some circumstances they won't give you an accurate read of whether you
are inside a large turnpoint cylinder or not.

Steve Kroop at Flytec <<info>>
writes:

I would like to add that, from a practical stand point, apparently
it is not realistic to measure distances in real time using the ellipsoid model
in a small handheld instrument (requires far too much processing power). I
discussed this with Erich and he told me that he was unable to find an algorithm
that could perform reasonably on even a fast mobile CPU (like smart phone and
tablets and like what is expected in the 7000). He looked into the current
flight apps available and they either used the FAI sphere or some weird
approximation of the WGS84 Ellipsoid.

Andrej Kohler at Naviter <<info>>
writes:

This is a good and the right solution.

Somehow the FAI sees SeeYou as a commercial operation and they don't invite us
to these meetings, which I do not mind at all. I really appreciate your efforts
in keeping us in the loop about all this.

SeeYou 6 has the selection box back in Tools > Options > General > Distance
calculation. We will remove this option from Oudie 3. It was introduced last
year during the European Paragliding championships in Bulgaria and it was a
nuisance to pilots, scorers, everyone.