8 hours ago, MattNischan said:

I don't believe so, we've been doing the full suite of 424 legs with all the various transition cases in all our avionics for several years now (starting with the NXi).

Last I looked we're roughly as compliant as the CL650, which is to say we're about as good as a real FMC. So, I'm not totally sure where their comparison is coming from. I believe the Fenix does quite well in the test suite also (and the FBW A32NX is starting to close the gap as well).

Exactly what I was thinking 

6 hours ago, MattNischan said:

I suppose at that point it really depends on your personal definition of "quite well". There are definitely real world boxes that get flown commercially daily that don't necessarily perfectly hit every case either, although like sim avionics, they get updated over time. That's a list of some of the least intuitive (and sometimes borderline/non-compliant on the spec in design) procedures. From a more global standpoint those particular cases are in the vast minority and their construction isn't necessarily well covered (or in some case at all covered) by spec. So, you have to compare the chart to the computations and continue to add more and more design specific edge cases.

It isn't really just a matter of loading up the latest PANS-OPS volume and coding to that (and especially not 424, which covers nearly zero of these types of pathological cases). In other words, saying one is coding to 424 and PANS-OPS doesn't indicate case completeness (it's what we're all doing already), as the cases that break and their resultant solutions aren't enumerated in those documents outright anyway.

Can confirm when flying out of Heraklion the once it drew some weird path.