Following some recent bad experiences with commercial simulators, and the prospect of simulating (from an FDM perspective) a modern jet in FlightGear, I am planning to work on some features which might eventually let me use FG for the stuff I enjoy flying.
As a starting point, there are a few infra-structure things I'm planning
to work on, and wanted to check for possible collisions. Especially with
upcoming auto-pilot and FMC work, I suspect.
Basically, I'm planning to start with flight plans / routes. This would
be a generic store for fragments of navigation path, not just 'real'
flight plans, with an associated XML format. I'd plan on using one basic
set of waypoint classes in various containers, as follows:
waypoint types: [and other things that might be in a route]
- airport (via IACO)
- fix (lon / lat)
- NDB / VOR
- intersection
- DME distance + radial from a VOR
This would be setup as FGWaypointBase and sub-classes. Others types
might be necessary like DME arcs, crossing-restriction altitudes and so
on.
The containers classes for these waypoints would be:
- flight plans
- airways
- SIDs / DPs
- STARs
Sample of the XML format:
<flightplan>
<head>
<aircraft=../>
<description>Edinburgh to Heathrow shuttle</description>
.. other stuff as required ..
</head>
<body>
<airport iaco="EGPH"/>
<sid id="TALLA-1H"/>
<!-- it's implicit that the next waypoint is some valid
transition from the DP -->
<vor ident="DCS"/>
<!-- ditto for the airway transition -->
<airway id="UB1" alt="FL290"/>
<vor ident="MGT"/>
<star id=" ..."/>
<airport iaco="EGLL"/>
</body>
</flightplan>
Implicit is that you can put an alt="6000" or alt="FL270" on most nodes,
and if that no alt is specified you'd use the previous one (with some
logic to make climbs and decents work. People who use VFR flight plans
might want to let me know if this will suffice, since I pretty much fly
IFR exclusively.
Other points of interest:
- there's lots of airway, SID and STAR data in various text formats on
the web, adding importers to these formats ought to be pretty easy. I
think Robin Peel's data may include Airway data already, need to check
that.
- Once a disk format is determined, it's possible to pester flight
planner authors to get support (eg from FSBuild or that Java one
designed for X-Plane..). Who knows, someone might even write a custom FG
planing tool.
Arrgh, this turned out much longer that expected. Anyway, I'm already
working on the XML and storage side of this stuff, but I would like some
agreement that this is a sensible approach before I commit to the design
too much.
[BTW, I'm intending to start with just simple flight plans, i.e only
airports / navs / fixes (what is already there) for the first cut,
obviously doing all the above is a decent amount of work..]
H&H
James
signature.asc
Description: This is a digitally signed message part
