On Thursday, December 31, 2015 at 1:16:08 PM UTC-8, wrote:

The often quoted manufacturers recommendation to not enable stealth is taken out of context. They are cautioning that leaving the Flarm in stealth when not at a contest will, of course, limit what is displayed on your Flarm and others. This designed feature is for use at contests not general flying.


Hey Sean,

I've read the spec over and over. I've read other Flarm-produced documentation and I've spoken at one point or another to most of the Flarm leadership team, including spending a day with Urs in his California office. Without putting specific words in anyone's mouth it was made pretty clear to me that the recommendation against using Stealth specifically includes contest use. They have said (in writing) that the only reason to use it (and implied that the only reason it was developed) was to forestall people turning their units off altogether. You can speculate that this was due to liability concerns, but no one ever said that to me and the way people talked about it just reinforced the sense that the main consideration is that they thought using it was a bad idea on technical grounds - second only to not using Flarm at all.

We can talk about the (small) probabilities of glider midairs and merits of a bit less situational awareness against the (still to be described in a quantifiable way) reduction in "tactical use", but it really is a line drawing exercise that makes many of us pretty uncomfortable - since the costs are some small increment in a horrific outcome versus a benefit that can't be adequately described, let alone quantified or demonstrated as to how it decreases the accuracy to which contest scores represent soaring skills.

Take John's confessional and unpack it a bit into the real implications. There are two main tactical use cases he describes:

Use Case 1: Knowing where other gliders are and to some extent how they are progressing - a different line, a different thermal, etc. Knowing where a handful of other pilots are some of the time - maybe where they made their turn in the cylinder - give you a sense of whether you are gaining or losing. In other words, it give you a sensation that you are racing - which a lot of pilots seem to enjoy. You have tactical information that you are, for instance, losing ground to another glider - instead of waiting until dinner time and the scores to figure this out. At three miles distance it gives you almost no useful information about what to do about it. Three miles laterally is a hard gap to close on a leg - you just give up too much to make a sharp deviation and if you do it with a gradual course change it's equally pointless as whatever is happening differently will likely be totally changed 10 miles or more down the course line. As prior data analysis - and experience - has demonstrated, following someone from 3 miles behind is generally a recipe for getting 4 miles behind unless you find your own, better thermals - that is, fly you own flight.

Use Case 2: Having some confidence under marginal conditions when there are other gliders about that you will have a decent shot at finding a thermal if there is one to be had. This reduces the chances that you will miss the saving climb that prevents a landout (or being stuck for a long time) when climbs (and particularly good climbs) are few and far between. From looking at the names on the list of landouts on days like this, it seems that there is more luck than skill involved, but it is possible that there is some skill, some local knowledge and some risk tolerance involved (e.g. willingness to drive down to 300', fly over unlandable terrain, into tight canyons, etc). Are these the types of soaring skills we want to value? Is having more landouts a desirable way to ensure these skills are tested? Does it really make the sport more attractive to new pilots to know we specifically want them to not have the information that could have gotten them home on a marginal day or when they are low and desperate? Landouts are the enemy of fair scoring - they scramble the scoresheet and we can't even come to a stable view of how to score miles versus miles per hour. The points we grant for landing out have come up and up over the years specifically because we realize landouts are mostly an indication of bad luck more than lack of skill. We devalue days with lots of landouts (luck factor). We've increased landout scores to the point that slow finishers are starting to complain that they don't get enough points. Why is it a good idea to deny pilots useful tactical information to avoid a landout? In addition to the basic fairness and values issues, it is a question the tort lawyers will be interested to address the next time we have a landout fatality in a contest - all the IGC files will be analyzed to see if there were any climbing gliders in range to avoid the tragedy. The next question will be "who is to blame for deliberately denying the pilot this potentially life-saving information". At that point what actually would have happened in the alternate case won't matter.

I wouldn't describe all opponents of Open Flarm as technophobes - some are (and have admitted to me that they are "not computer people"), other like technology just fine but seem to feel that some skill they have (perhaps risk-tolerance is one) will be diluted with new and better information. But denying people all external information hardly seems like a fundamental principle of glider racing - if it were we would do separate time trials or all MAT format to maximize the separation of gliders so you can't use any visual cues. We would have leeching penalties that are quite easy to calculate with IGC files. We don't, and no one seems to be interested in going down that path.

9B