Suggestion: HDOP threshold setting

[W-Geronius]

Some GPS receivers report a 3D fix despite very poor DOP values. It might be better not to rely too closely on these positional data.
What about a HDOP threshold to invalidate a poor fix? Acceptable ranges may be documented as in http://www.noomio.com.au/wiki/faqs/gnss-setting-threshold-for-acceptable-accuracy/

Example:
sensor ATGM336H as included in OBP60 V2 reported a fix at HDOP > 25.0, resulting in position jumps within several hundreds of meters.

[wellenvogel]

This functionality heavily depends on the available NMEA0183 data.
Some records like RMC do not report HDOP.
So it at least requires GSA.
And it could be difficult to decide if a HDOP reported by GSA is relevant for a position reported by RMC.

[W-Geronius]

Sure, no DOP when GSA is missing - in that case, as well as when RMC talkerid differs from GSA talkerid, no threshold can be applied