Bigsnut said..
It's all over for JulienLe as soon as a new device is found. I don't know how everyone is being so tolerant and sucking up because it's the only device available when he wont deliver.
You are describing what pretty much has happened with GPS-speedsurfing.com. They have worked with at least 3 companies to make a speedsurfing GPS, including Julien. That involved considerable effort, and at least in one case also payments to the GPS developers. But 2 of the 3 efforts ended with the GPS3 folks being disappointed. They seem to be reasonably happy with the result of their collaboration with Coros, though. The price for the GPS3-approved Coros Apex Pro and Apex Pro is around $500 US, quite a bit higher than for Motions. But it is readily available.
The chances that "a new device is found" to replace the Motion are very close to zero. There is no
clearly defined process to get a new device approved, other than "send some files and get and ok from sailquik". In contrast, the GPS3 web site has clearly defined criteria. They are described for "official GPS Record devices" (
www.gps-speedsurfing.com/default.aspx?mnu=item&item=gpsother), but the relevance extends beyond records, which are of interest for only a few.
The very ironic thing about this is that one of the criteria that keeps other devices from being approved is the requirement for error accuracy data. While that's a great idea, those accuracy data are almost never actually used by the analysis software to identify artifacts that other filters would have missed. For example, ka72.com appears to use a filter threshold that GT-31 data can
never reach. For Motion data, the error estimate thresholds in all analysis programs are so high that they completely miss many artifacts. Here's one example of a 3 knot spike that the filters don't catch:

So the situation is that many available GPS devices will not even be
considered for the GPSTC because they don't give error estimates, while the error estimates of the "approved" devices are basically
not used for a very large fraction of the GPSTC postings.
So back to Julien's question, specifically in context of the GPSTC:
- specify
exactly and
publicly what what would be needed to get another device approved
- replace the error estimate requirement with criteria for
actual errors