What features would you like to see in P3Dv5?

[Dominique_K 10,149]

I've skipped v4  and I'll go v5 if LM  does something for at least one of these issues :

1- Terrain

 - No repositioning of default (ie outside OrbX fat regions) airports which are for the vast majority somewhat off with the consequence, for instance, of roads, tracks, streams crossing many airport areas

 - No sloping runways which leads to runways partially or entirely in trenches or plateaus when you add a sharper mesh. . 

Tired to do the job myself with ADE (whenever possible).

2- Aircraft

Water physics for floatplanes is arcadish

 3- Atmospheric light rendering engine outdated. 

4- FS2K2 old ATC 

 

 

 

 

[GSalden 4,657]

1 minute ago, Ray Proudfoot said:

I don’t see the need for a brand new core engine for P3D. The biggest improvement came with 64-bit. How soon we accept that as the norm and continue to demand more.

Many of us have invested serious money in 3rd party addons. Are you seriously wanting all those to be junked and purchased again? And what of the developers who try to sell these new versions? How great would the uptake be? It could mean some of them don’t make a profit and fold.

What I would like to see are the flaws in the core ESP code corrected. Ai have very little intelligence. That needs improving so they don’t taxi straight into you or line up on the runway when you’re about to land.

Stars appearing with the sun above the horizon should be an easy one to correct so why has it been ignored for so long?

Does L-M Have authority to change this code? Perhaps not which is why the flaws remain.

Speaking personally I’m more than happy with how P3D has evolved since L-M took over. They saved flight sim for many of us and I’m eternally grateful for that. Without them we’d still be flying FSX which is a truly horrifying thought.

Imho we really do need a new engine. No more a 2004-2005 based engine..

Use a multi view setup and see how the framerate drops. 

 

[Ray Proudfoot 8,955]

11 minutes ago, GSalden said:

Imho we really do need a new engine. No more a 2004-2005 based engine..

Use a multi view setup and see how the framerate drops.

How many people have such a setup? A minority. And are you prepared to pay hundreds of (your currrency) for that with all that is involved?

[simbol 5,819]

25 minutes ago, Ray Proudfoot said:

Does L-M Have authority to change this code? Perhaps not which is why the flaws remain.

They have been changing the ESP code since day one...

S.

[GSalden 4,657]

12 minutes ago, Ray Proudfoot said:

How many people have such a setup? A minority. And are you prepared to pay hundreds of (your currrency) for that with all that is involved?

Yes, for a new engine where the main thread is split up over multiple cores , so the framerate will go up and the Sim will be much smoother , I will be happy to pay for that.

It is like driving the latest model car with an engine that is 15 years old ...

Look at what high end pc’s we need if we want eyecandy + smoothness.

 

 

Edited March 11, 2019 by GSalden

[Ray Proudfoot 8,955]

43 minutes ago, simbol said:

They have been changing the ESP code since day one...

S.

Then why do these long-term problems not get resolved?

42 minutes ago, GSalden said:

Yes, for a new engine where the main thread is split up over multiple cores , so the framerate will go up and the Sim will be much smoother , I will be happy to pay for that.

It is like driving the latest model car with an engine that is 15 years old ...

Look at what high end pc’s we need if we want eyecandy + smoothness.

Maybe I'm missing something but I'm already seeing all 12 of my cores being utilised.

[Dazkent 113]

full screen with vsync support, been asking for that since V1 haha 😁

[simbol 5,819]

2 hours ago, Ray Proudfoot said:

Then why do these long-term problems not get resolved?

The LM Roadmap is their own business and I cannot comment on such.

However, when I joined the LM Beta team (end of October 2018) I found out that lots of problems / bugs that were bothering me as a developer in terms of effects were very well know by other developers and yet these were never reported to LM -reasons unknown, and this is beyond to my comprehension -, since then I have been reporting any anomaly and problems I find, and so far every single bug reported has been acknowledged by Lockheed Martin engineers.

So I will throw a guess answer to your question:

A) Because all we do as users is moan on these forums instead of raising our voice and place either a formal bug report or feature request directly to Lockheed Martin engineers using their official forums / channels.

B) Because many developers (including freeware devs) in the flying sim community have failed to request or report issues that bother many users in the flying sim community, this is due to many different causes, they are too busy with their own problems, they only test their own add-on's and nothing else, they are disconnected with the flying sim community as they have their own forums and therefore they only have a limited view of the global user requirements, etc.

On the plus side, you have also other developers that actually do report things and ask for new features, @Lorby_SI, is a very good example as he always takes the time to post any issues he finds with the simulator or the P3D SDK, I know he can feel sometimes he is being ignored, but as an insider of the LM beta program I can tell this is not the case .

Conclusion

Don't waste your time writing on public forums, Facebook, Reddit, Twitch, etc. things that you would wish LM engineers to implement, go to their official LM forums and ask them directly as you will have much bigger chances to being listened to, and a word of advise, just because they don't respond to your request it doesn't mean they haven't taken note of your request / suggestions.

Kind Regards,
Simbol

 

 

Edited March 11, 2019 by simbol

[DellyPilot 109]

29 minutes ago, simbol said:

go to their official LM forums and ask them directly as you will have much bigger chances to being listened to, and a word of advise, just because they don't respond to your request it doesn't mean they haven't taken note of your request / suggestions.

Great to hear simbol, I posted my 2 cents last week here (My 3 big wishes for P3D v5)  https://prepar3d.com/forum/viewtopic.php?f=6317&t=132748 and linked back to this thread.

Hopefully someone in LMWorld takes a little look. I guess the issue is that posting on their forums is unlikely to have as big an audience as here and you will rarely see a post from LM. Even though I agree with you people just think whats the point.

[DellyPilot 109]

1 hour ago, Ray Proudfoot said:

Maybe I'm missing something but I'm already seeing all 12 of my cores being utilised.

It absolutely uses all cores (16 over here!) but importantly not for all jobs. Some compute jobs parallelise nicely like loading and processing autogen, but some (rendering related where synchronisation is key) are only single threaded, thus creating bottlenecks.

Elegant engine design is about chasing bottlenecks. There will always be one limiting factor, but ideally faster hardware should = better performance. P3D has to run on millions of possible hardware configs and remain backward compatible, what they have done really is a brilliant feat of engineering but it is now not keeping up with hardware.

I see 100FPS but stutters, 40FPS but wait 6 minutes for terrain to load when I arrive at my destination. This is so fixable.

 As a programmer I want to know more about the P3D engine architecture, its a powerful beast but has these issues that lie right at the core of the engine. 

I'd really love LM to do a developer blog/vlog on the engine design, I think building more of a community would actually provide them with some great ideas and input.

Edited March 11, 2019 by DellyPilot

[Ray Proudfoot 8,955]

7 minutes ago, DellyPilot said:

It absolutely uses all cores (16 over here!) but importantly not for all jobs. Some compute jobs parallelise nicely like loading and processing autogen, but some (rendering related where synchronisation is key) are only single threaded, thus creating bottlenecks.

Elegant engine design is about chasing bottlenecks. There will always be one limiting factor, but ideally faster hardware should = better performance. P3D has to run on millions of possible hardware configs and remain backward compatible, what they have done really is a brilliant feat of engineering but it is now not keeping up with hardware.

I see 100FPS but stutters, 40FPS but wait 6 minutes for terrain to load when I arrive at my destination. This is so fixable.

 As a programmer I want to know more about the P3D engine architecture, its a powerful beast but has these issues that lie right at the core of the engine. 

I'd really love LM to do a developer blog/vlog on the engine design, I think building more of a community would actually provide them with some great ideas and input.

I imagine the refinements on processing via logical cores will continue. The horse (processors) has to come before the cart (P3D).

1 hour ago, simbol said:

Don't waste your time writing on public forums, Facebook, Reddit, Twitch, etc. things that you would wish LM engineers to implement, go to their official LM forums and ask them directly as you will have much bigger chances to being listened to, and a word of advise, just because they don't respond to your request it doesn't mean they haven't taken note of your request / suggestions.

An excellent post, well done. It had crossed my mind on more than one occasion to post about these problems on their forums. I shall now do so.

[DellyPilot 109]

5 minutes ago, Ray Proudfoot said:

I imagine the refinements on processing via logical cores will continue. The horse (processors) has to come before the cart (P3D).

I like the pun even though I don't fully get the point 😂

Edited March 11, 2019 by DellyPilot

[Ray Proudfoot 8,955]

23 minutes ago, DellyPilot said:

I like the pun even though I don't fully get the point 😂

What I was trying to say was the hardware has to be available before the software can take advantage of it.

[eslader 1,783]

Funnily enough, that wasn't really MS's philosophy. FS-4 was the last sim to run smoothly with all the graphics options turned up on hardware available at the time of release. From then on, you had to wait a number of years from Flight Simulator release to the point where you could get your hands on hardware that could run it smoothly. And if you decided to get high-test addons like weather engines or detailed scenery, all bets were off.

 

[Airline2Sim 489]

Aside from the obvious (performance improvements, better rendering etc) for me it would be nice to have some of the things that X-Plane does rather well, such as sloping taxiways and runways, better AI road traffic and lighting. 

Aside from that, the ability to control runways in use directly from the sim and highest up on my wish list would be a complete rebuild of the ATC and AI system. The default ATC is hopelessly off the pace and AI behaviour is at best erratic and at worst plain daft. 

If the base ATC and AI platform was improved then I’m sure add-on developers would pile in to customise it. The integrated ATC is hopeless in many ways but it’s still the slickest and easiest to use out of all the options, which have to be built around the sim.