This PR refactors the Doctrine class to load from YAML files in the
resources folder instead of being hardcoded as a step towards making
doctrines moddable (Issue #829).
I haven't added anything to the changelog as a couple of things should
get cleaned up first:
- As far as I can tell, the flags in the Doctrine class (cap, cas, sead
etc.) aren't used anywhere. Need to test further, and if they're truly
not used, will remove them.
- Probably need to update the Wiki
1000km isn't large enough in the case where there's an off-map spawn
that's a long way from the target, but still in range for aircraft like
the B-1. Double it, which for now is enough to fix the one pathological
case we know.
Fixes https://github.com/dcs-liberation/dcs_liberation/issues/3156.
Create a WaypointAction class that defines the actions taken at a
waypoint. These will often map one-to-one with DCS waypoint actions but
can also be higher level and generate multiple actions. Once everything
has migrated all waypoint-type-specific behaviors of
PydcsWaypointBuilder will be gone, and it'll be easier to keep the sim
behaviors in sync with the mission generator behaviors.
For now only hold has been migrated. This is actually probably the most
complicated action we have (starting with this may have been a mistake,
but it did find all the rough edges quickly) since it affects waypoint
timings and flight position during simulation. That part isn't handled
as neatly as I'd like because the FlightState still has to special case
LOITER points to avoid simulating the wrong waypoint position. At some
point we should probably start tracking real positions in FlightState,
and when we do that will be solved.
This fuzz test generates random inputs for waypoint solvers to check if
they can find a solution. If they can't, the debug info for the solver
is dumped to the testcases directory. Another test loads those test
cases, creates a solver from them, and checks that a solution is found.
Obviously it won't be immediately, but it's a starting point for fixing
the bug and serves as a regression test afterward.
This is a replacement for the existing "zone geometry" classes that are
currently used for choosing locations for IP, hold, and join points.
The older approach required the author to define the methods for
choosing locations at a rather low level using shapely APIs to merge or
mask geometries. Debug UIs had to be defined manually which was a great
deal of work. Worse, those debug UIs were only useable for *successful*
waypoint placement. If there was a bug in the solver (which was pretty
much unavoidable during development or tuning), it wasn't possible to
use the debug UI.
This new system adds a (very simple) geometric constraint solver to
allow the author to describe the requirements for a waypoint at a high
level. Each waypoint type will define a waypoint solver that defines one
or more waypoint strategies which will be tried in order. For example,
the IP solver might have the following strategies:
1. Safe IP
2. Threat tolerant IP
3. Unsafe IP
4. Safe backtracking IP
5. Unsafe backtracking IP
We prefer those in the order defined, but the preferred strategies won't
always have a valid solution. When that happens, the next one is tried.
The strategies define the constraints for the waypoint location. For
example, the safe IP strategy could be defined as (in pseudo code):
* At least 5 NM away from the departure airfield
* Not farther from the departure airfield than the target is
* Within 10 NM and 45 NM of the target (doctrine dependent)
* Safe
* Within the permissible region, select the point nearest the departure
airfield
When a solver fails to find a solution using any strategy, debug
information is automatically written in a GeoJSON format which can be
viewed on geojson.io.
Fixes https://github.com/dcs-liberation/dcs_liberation/issues/3085.
This looks like it was just a typo. We want to join as late as possible
to allow flights coming from other airfields to take the best route to
the target that is safe, rather than joining as early as possible, which
isn't useful since pre-join and post-split are supposed to be safe areas
anyway.
This is briefly moving us over to my fork of pydcs while we wait for
https://github.com/pydcs/dcs/pull/206 to be merged. The adaptation is
invasive enough that I don't want it lingering for long.
* Initial refueling definitions.
* Adding refuel definitions.
* Initial functionality changes
* Regenerate package when adding refueling flight.
* Recursively change package waypoint.
* Fix mypy errors.
* Regenerate flight plans when tanker is added to package.
* Give tanker better starting position on package recovery.
* Add TOT calculation for refueling waypoint.
* Timing changes to Strike split point and Refueling start time.
* Add correct waypoint builder for refuel in tarcap and sweep. Remove restrict afterburner on refuel point.
* Always generate a refuel point for a package.
* Less arbitrary altitude in Refuel track start time calculation.
* Refueling waypoint no longer optional.
* Fix mypy gen error.
* Better discrimination of which tanker flight plan to make.
* Remove refuel tot calculations.
* Remove package regeneration on tanker flight addition.
It's rare with the current 5NM buffer around the origin, but if we use
the hold distance as the buffer like we maybe should it's possible for
the preferred join locations to fall entirely within the home zone. In
that case, fall back to a location within the max-turn-zone that's
outside the home zone and is nearest the IP.
Test cases:
1. Target is not threatened.
The IP should be placed on a direct heading from the origin to the
target at the max ingress distance, or very near the origin airfield
if the airfield is closer to the target than the IP distance.
2. Unthreatened home zone, max IP between origin and target, safe
locations available for IP.
The IP should be placed in LAR at the closest point to home.
3. Unthreatened home zone, origin within LAR, safe locations available
for IP.
The IP should be placed near the origin airfield to prevent
backtracking more than needed.
4. Unthreatened home zone, origin entirely nearer the target than LAR,
safe locations available for IP.
The IP should be placed in LAR as close as possible to the origin.
5. Threatened home zone, safe locations available for IP.
The IP should be placed in LAR as close as possible to the origin.
6. No safe IP.
The IP should be placed in LAR at the point nearest the threat
boundary.
