Re: The endless influx of low-quality, low-realism, quantity-over-quality aircraft into FGAddon

How do you explain setting the sliders to the max only produces 10 
aircraft if there are so many incorrectly rated aircraft? Maybe at the 
mid range settings there is more contention as to the accuracy of the 
rating. If I changed anything, maybe I would add an FDM switch that 
could separate the FDM's so one could get ratings for JSBSim 
FDM's/aircraft and one could get ratings for YASsim FDM's/aircraft. I 
think that is most of the complaint. That way you won't alienate those 
that like the YASim FDM's and give them a penalty simply for preferring 
one FDM over another.



On 5/22/2026 11:35 AM, Josh Davidson wrote:
> Ultimately, I'm less concerned with restricting FGAddon, and more of 
> the inaccuracy of aircraft rating, which is what causes "poor user 
> experience" but them seeing an aircraft rated too highly, and being 
> disappointed, and blaming the simulator (a valid point). Correcting 
> ratings or making sure aircraft in FGAddon are being rated accurately 
> mostly solves this problem.
>
> I will explain lower in this email in a reply to Federico why based on 
> the existing FDM guidelines, a YASim aircraft cannot score above a "2" 
> - but this is violated by "TONS" of aircraft in FGAddon currently. 
> Spoiler: It is not possible to "tune rate of climb and cruise for POH 
> performance numbers" in YASim - or if it is, I am not aware of "ANY" 
> aircraft that currently achieves this - please point me to one if you 
> know of one.
>
> From the wiki: 
> https://wiki.flightgear.org/Aircraft_rating_system#Flight_Dynamics_Model_(FDM) 
> <https://wiki.flightgear.org/Aircraft_rating_system#Flight_Dynamics_Model_(FDM)>
> *Flight Dynamics Model (FDM)*
>
>   * 0: None, or using FDM from other aircraft
>   * 1: JSBSim Aeromatic or YASim geometric model used without tuning.
>     Flaps modeled.
>   * 2: FDM tuned for cruise configuration.
>   * 3: FDM tuned for rate of climb and cruise Pilot Operating Handbook
>     (PoH) performance numbers
>   * 4: FDM matches PoH in 90% of configurations
>   * 5: FDM matches PoH and most known test data.
>
>
> Lawrence:
> > And what if, Josh want to develop the B738 (modern version),
> and time later he opt to also develop the old B738, (just for the sack
> of the beauty of some non glass-cockpit) ?
> Naming them both B738 or B738-Josh also do not solve the problem. Just
> saying.
>
> I didn't understand that. The B738 is only glass.
>
> If I wanted to make an aircraft like the B737-300 which had both 
> classic non-glass and partial-glass, I can do that in the same package 
> - see for example the MD-80 which does exactly that. There is an 
> option to equip either analog classic gauges or EFIS. No separate 
> package is needed.
>
> This is not the name collision I spoke of. For example, if Josh (me) 
> and FrankVMD develop the 777, then he already has his as 777-VMD I 
> believe, but then there is the standard FG 777 (of which I am 
> contributing a new flight model and improved systems), which is still 
> named "777". I guess that is ok in that regard, given that it is an 
> update to that existing package. But perhaps going forward, it may be 
> advisable to have the package name include the author or at least the 
> primary author/team in cases where more than one developer is working 
> on it.
>
> > As we can read, it is clearly not possible to look at "quality" using us
> all the same criteria.
> Even the same person might lookup, for a few flights only, an aircraft
> with different criteria...
>
> I don't agree - the star system already in use in FG rates "quality" 
> based on "fidelity". Essentially, the higher the stars, the more 
> accurate/authentic the simulation is (for systems/FDM) or the more 
> graphically advanced (for Model/Cockpit).
>
> There is no need to change this, only to correct the incorrect rating 
> of certain aircraft who are rated much higher than they should be - 
> especially in undermaintained aircraft which might still be evaluating 
> visual rating based on 15 years ago or more quality. But in the 
> systems/FDM department, the scale hasn't shifted considerably in 
> either the commercial sims or here - but until the last 10 years or 
> so, highly detailed systems and such were not really part of FG.
>
> Federico:
> >  I'm not sure why people say the FDM isn't good on them, I've loaded 
> several of his aircraft (just random ones) into my FDM editor and they 
> look great.
>
> The shape of the elements per FDM validity is only a marker of 
> accuracy when used with computational fluid dynamics (CFD). YASim code 
> essentially takes the input shape data and uses it to modify some 
> pre-determined aero behavior. There is "NO" analysis/CFD/etc occurring 
> here - as such things "CANNOT" be done at runtime due to limitations 
> in processing power. A single high-fidelity CFD run takes much longer 
> than a single FDM cycle (60Hz).
>
> I have flown a few types of planes in real life, all GA - and am 
> fairly familiar with how larger aircraft types behave due to my own 
> simulation and aerodynamics work. I have never *ONCE* encountered a 
> YASim FDM that flies like a real plane OR follows POH numbers to any 
> reasonable closeness. I welcome anyone to point me to one and I will 
> try it and report back. Many of Emmanual's planes I tried have major 
> issues such as inability to climb and in some cases, even hold 
> altitude, without losing speed, with full throttle when the flaps are 
> extended to the nearly full or full positions, amongst others, aside 
> from the "standard" YASim problems of inaccurate engine simulations, 
> incorrect stall dynamics, wrong lift/drag curve behavior as alpha 
> varies, etc, making them fly very strangely.
>
> I created a mailing list thread some time ago asking about some of 
> these YASim problems as the problem seems to be with YASim itself, 
> given that it gives practically "ZERO" control over the flight 
> dynamics to the developer, aside from deciding the shape of the 
> geometry, and the effectiveness of flight control surfaces (by setting 
> their lift/drag offsets).
>
> Emmanual has stated, incorrectly, that YASim evaluates the shape and 
> computes aerodynamic qualities based on the shape of the aircraft, and 
> thus he tries to match the geometry in YASim to the actual shape of 
> the plane and assumes, again incorrectly, that the plane flies 
> "properly". I have found in order to reduce some of the YASim 
> "funnies" you actually have to make the geometry somewhat incorrect - 
> because at the end of the day, it's not doing any type of 
> CFD/aerodynamic evaluations, but rather simply applying modifications 
> to pre-determined behavior.
>
> If you doubt this analysis, please take a look at the source files 
> themselves:
> https://gitlab.com/flightgear/flightgear/-/tree/next/src/FDM/YASim 
> <https://gitlab.com/flightgear/flightgear/-/tree/next/src/FDM/YASim>
>
> Kind Regards,
>
> --
> Josh Davidson
> ------------------------------------------------------------------------
> *From:* Alan Teeder <[email protected]>
> *Sent:* Friday, May 22, 2026 10:49
> *To:* [email protected] 
> <[email protected]>
> *Subject:* Re: [Flightgear-devel] The endless influx of low-quality, 
> low-realism, quantity-over-quality aircraft into FGAddon
>
> Sorry, but for many users Flightgear *IS* a game.
>
> For me Flightgear is a flight simulator, but I consider myself 
> probably in a minority.
>
> Alan
>
> On 22/05/2026 16:36, David Hudach wrote:
>> Personally, this is inappropriate for this discussion: */Nice troll 
>> though ;)/*
>>
>>
>> Flightgear-devel mailing list
>> [email protected] <mailto:[email protected]>
>> https://lists.sourceforge.net/lists/listinfo/flightgear-devel <https://lists.sourceforge.net/lists/listinfo/flightgear-devel>
>
>
> _______________________________________________
> Flightgear-devel mailing list
> [email protected]
> https://lists.sourceforge.net/lists/listinfo/flightgear-devel

_______________________________________________
Flightgear-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/flightgear-devel