I did some experiments with FPL files and a Garmin 695. When importing database waypoints all the 695 cares about is the name and the country code of the waypoint. It then searched the internal database and sources the coordinates there. Only if you have user waypoints in the FPL file it will use the coordinates. I wrote a gpx2fpl converter. And the difficult part is the country code. Most other formats do not contain this information. You have to use some sort of vector file with the FIR boundaries. If you do not all waypoints existing multiple times will be rejected. For example a "SL" NDB does exist about 10 times worldwide...
I would assume it uses the Garmin coordinate format internally. The Flight Plan Migrator kits converts the XML to another representation which it writes onto the data card.
LAT/LON = Garmin format * 90 / (1 << 30);
That's a horrible system, which more or less guarantees periodic mismatches.
I suppose the software has to treat the floating point numbers as textual strings, not as floats, otherwise (short of using BCD floats) the chance of ending up with the same one millionths of a degree is just about zero.
As I say, if you have a current GPS DB, and you have a current validated route (not one developed a year ago), you should be able to simply spell every waypoint and it should work.
The main advantage of VF according to their presentation is that you can use airways which the GNS do not know about. This obviously requires a correct and up to date database in VF.
If one just used waypoints, then it could be theoretically implemented without any database unless the GNS crossfill protocol requires coordinates which it most likely does.
This is how an .fpl file looks like that you upload to the GNS using a special card and the Garmin Flight Plan Migrator Kit:
How could it guarantee correct functioning with an outdated database?
It takes care of itself, because
the Eurocontrol-validated route will be wholly on current-AIRAC-cycle waypoints, as should the resulting plog
one should have a current DB in the GPS (OK I know many don't, and sometimes I forget to update, but then all bets are off)
As I say, if you have a current GPS DB, and you have a current validated route (not one developed a year ago), you should be able to simply spell every waypoint and it should work.
More likely, as with the XML flight plan format, the GNS needs coordinates together with the waypoint name. If those coordinates don't exactly match what it has in the database, you end up with a locked waypoint in the flight plan. While this works, it's a bit of a nuisance.
Hmmm... sounds likely. I recall this issue with loading flight plans into the old Skymap 2 
There must be more to this crossfill protocol than meets the eye. It probably isn't a simple case of sending a list of waypoints.
More likely, as with the XML flight plan format, the GNS needs coordinates together with the waypoint name. If those coordinates don't exactly match what it has in the database, you end up with a locked waypoint in the flight plan. While this works, it's a bit of a nuisance.
OK; that's outrageous.
No, it's logical. How could it guarantee correct functioning with an outdated database? The best it can do is warn you about the mismatch and still let you continue. This is exactly what the product does according to the installation manual.
It also suggests that they had a problem reverse engineering the "both units need the same database version to crossfill" feature.
Where did you get that? The manual states that a VF with a database older than the GNS will work but of course cannot be guaranteed to work 100% due to database differences.
OK; that's outrageous. That means one has to keep both updated, or keep both equally out of date.
It also suggests that they had a problem reverse engineering the "both units need the same database version to crossfill" feature. That is actually really interesting. It might indicate that one unit transmits a hash of the database version to the other unit, and VF could not emulate that.
Otherwise, why the hell can't one just spell the waypoints - as the video shows?
There must be more to this crossfill protocol than meets the eye. It probably isn't a simple case of sending a list of waypoints.
The installation manual section 6.7 says that if the databases don't match, there will be a warning when you enter a flight plan and some waypoints might not work correctly.
I've been asking questions on a US site and this is the current reply
As of yet, we have not established any VoiceFlight dealers over in Europe. Before selling any units over there, we would want to make sure that they could be properly supported by knowledgeable dealers.
There are two obstacles to accumulating a enough European customers to establish a viable dealer network. One is the lack of the European certifications, and the other is the cost of the VFS101 Navigation updates that would be required.
Quite frankly, at the moment, we are scrambling just to keep inventory on the shelves and filling the dealer orders. But when the dust settles, our first step in this area will be to obtain the Canadian certifications and expand our dealer network to the North. After that, we will take a look at other locales.
which suggests a lack of understanding of aircraft certification; specifically that the world is full of N-regs...
The questions regarding the $25/month, whether one can update the database less often, etc, were left unanswered.
