LM implemented features hardly used or fail - But why?

[kevinfirth 1,392]

3 hours ago, G-RFRY said:

We should also ask why Devs Patched airports for v5 from v4 but don`t bother implementing the features of the full SDK. 

Because it was a) the easiest option with the least work; and b) the work required was too difficult or time consuming to justify on the ROI. This is where LM need to step up and overcome the difficulty/time barriers to make the ROI more attractive, otherwise Devs will always choose option a...

[jeansy 338]

2 hours ago, mpo910 said:

You may be right on this.

But don´t making developing for P3D more attractive is also not really motivating devs to bring nice enhancements to the sim. And that is something I really would like to see, also in future.

And if P3D and MSFS share lots of "tools" and "ways" to work, so the outcome is high and the dev speed is high too, but it takes less effort, then I think we all have some benefit from this.

- Devs can concentrate on more than 1 platform (risk spreading too)

- Devs can have faster ROI

- Devs can have more profit

- Users can have more addons for decent prices

- All have less risk being dependent on 1 single platform

This is what I would like to know if this could be possible.

Marcus

Reading through the 4 pages thus far, people are forgetting the P3D is a by product from military and civil simulator contracts which was never going to have the eye candy or features MSFS has for the home or upcoming console user, and more likely never will.

It is the home user that has the expectation of making it something it was never meant to be. The driving factors behind LM and what they produce and release is what the big dollar contracts ask them to supply. They may listen to the general public however for the past several yrs most of the "home user wants" have fallen on deaf ears.

LM has no interest in what MSFS or dollar compassion just a one or 2 contracts alone will earn LM more income than MSFS sales alone.

For now there is enough source codes and data to tailor your install to what you want. Over time maybe P3D will have future enhancements or new graphics engine that may resolve a lot of current gripes or maybe not.

As for the payware developers they are only going to follow the interest lies and where they can make quick money.  It might be a far left call, but I think a lot will continue with P3D once the hype of MSFS settles

 

Edited January 4, 2021 by jeansy

[kevinfirth 1,392]

29 minutes ago, Christopher Low said:

The only reason that I would upgrade to P3Dv5 is if the "big" stutters that I experience in quite a few locations across the UK would be considerably reduced (or removed altogether). However, the VRAM issue is a concern. If I could not run P3Dv5 with all of my current addons (and at the same scenery complexity and texture resolution settings) without suffering a CTD, then it is a non starter. I have a 6GB GTX 980Ti, and I run P3Dv4 @ 1080p resolution with 2048 textures. Reducing those textures to 1024 is not an option for me (I have seen the difference on my current PC).

Chris, so buy v5, try it out and if it doesn't work on your hardware, get a refund.

I found that going from an i7 8700K to an i9 10850K together with moving to v5 meant much less stutters if not entirely eliminating them. 

You have I think a pretty decent 8600 right? I run my sim at 4k, and I wouldn't want to try with anything less than 11Gb vRAM, but at 1080p I would hazard a guess that your 980ti with 6Gb will manage ok.

There's only one way to find out, test it 🙂

[Wolkenschreck 269]

3 hours ago, G-RFRY said:

I have no concern of the future survival of P3D i will bet you now the LM will continue with P3D long after MS stops putting money into MSFS which they will when finished and final. They have no plans for a new version only this for XBOX.

From a user perspective I totally agree with you. We got a fairly advanced sim with hundreds of add-ons that are available for us to use. And I, for my part, invested my Christmas budget solely on P3D add-ons.

The comment about survival of the platform was from a development perspective. Or to phrase it more precise: Survival in the sense of keeping P3D relevant for developers to produce content for it. And the plain truth is that P3D simply needs to modernize the tool set for development.

I do not know what Microsoft plans for the future. But expecting them to close the shop after the XBOX version release is just as likely as LM canceling the P3D program. Both could or could not happen. We simply do not know it. And than there is also XPL.

But for the foreseeable future is MSFS here to stay. Thus, what should a developer do? Continuing to put a lot of effort in a outdated development toolkit for P3D first or develop first with a more modern toolkit for MSFS (and with a bigger market at the moment) and see what assets can be reused for a later P3D conversion? If my monthly income would depend on this than the answer for me would be clear, as time is money.

Not catching up here will not make P3D irrelevant for us as users. But in terms of content it will be left behind. Just because we all love P3D should not make us blind eyed or defensive for against the changes happening in the sim world around us. Let's see if the following generation is keen on digging into the development environment of P3D. I simply doubt it. And do stress it a last time: It is not P3D that is outdated or the SDK functions. The features stated by Marcus are a proof that P3D is continuously modernized. It's just the toolkit for development that that needs some modernization.

Edited January 4, 2021 by Wolkenschreck

[emko 15]

2 hours ago, jeansy said:

It might be a far left call, but I think a lot will continue with P3D once the hype of MSFS settles

It looks like I am a far left-itst as well. Let's see where MSFS is going but I'm a bit skeptical. So far I'm yet to see a good quality add-ons which is what P3D makes interesting for me. Not really the case for X-Plane. I used to run XPL up to version 10.30. Both the sim and its add-ons were littered with half-baked features, every minor update was constantly breaking aircraft and add-on devs usually released aircraft in very unfinished state. Hopefully it changed.

 

From my extremely limited RW experience (a guy took me flying and let me perform a couple of turns in his aircraft) both light GA aircraft and LSA (Zlin 43, Alto 912) felt very stable to me in flight. I have very similar feeling in FSX and P3D (yes, the aircraft really felt like they were "on rails" - you feel what the aircraft is doing "by your a$$" and you subconsciously compensate for it - much like riding a bicycle or driving a car). XPL might be better at simulating physics of the flight however when you have no feeling in your controls and your chair it makes it needlessly hard. I prefer a simulator which simulates the way the aircraft behaves and feels rather than the way how it flies when you have extremely limited ways to compensate for it at home.

[W2DR 1,859]

I just have to say again.....Ford is better than Chevy, Ferrari is better than Lamborghini, and Sofia is better looking than Gina. It's all a matter of preference.

[d.tsakiris 417]

26 minutes ago, emko said:

yes, the aircraft really felt like they were "on rails"

A little off the topic, but this is also my experience

[dave2013 1,123]

12 hours ago, Benjamin J said:

However, the problem is not what you do in ADE, because I don't think anybody is really looking forward to a default style ground poly. The real trouble is making a custom ground poly that properly follows the ground, for which there are no tools at all, for as far as I can see. But even if you restrict yourself to the ADE alone, setting up the various height differences is a tedious and laborious process, especially for larger airports, that then also require significant retooling on the mesh front. All the while in MSFS the groundpoly that you design adapts to the existing mesh, and is thus a thousand times easier to work with. This, I think, is the way P3D should be going. All combined, it's just not a particularly enticing proposition to do sloped runways in P3D. In my own projects I've seriously considered doing it, specially for one of them where it could really benefit the airport scenery. It's also a rather small airports, so wouldn't require a ton of work. However, the more it became clear the amount of work that would need to go into it, I held off on it. Of course, that's just me. I consider myself a hobbyist - as I said before, I'm not in it for me. I'm mroe cocnerned about the time I have to work on these things.

I agree.  The airport background poly should also conform to the underlying terrain, as well as runways, taxiways, aprons, etc.

There should also be an automated mesh smoothing incorporated into the process, although  this may not be necessary for lower mesh resolutions.  It could be an option.

I recently made an airport for Calama, Chile and it looks awful because of the considerable E-W slope of the terrain - one side either sits atop a very high cliff which looks horrible, or I lower the cliff and end up with a terrain wall at the other end.

Hopefully LM will work on this as it is very important for airport realism.

Dave 

[Benjamin J 751]

8 hours ago, Wolkenschreck said:

MSFS is doing a better job here and makes development easier for newbies. Hence, you will see a far larger output of airports from new developers who just getting started with sim add-on design.

I agree with this entirely. Indeed, P3D's often-requested features seem to often go unused because their implementation can be vague, incomplete, confusing and/or prone to change.

Personally I had one huge misconception when I started developing scenery for P3D: "the modeling is the hardest part"

Now, modeling isn't easy, and it takes time to get to grips with the software (using the camera in a 3D modeling program for the first time, anyone?). I personally use Blender. But when you do get the hang of it, it's actually not too difficult and it becomes a straightforward process. The main thing to learn is the best strategy for modeling, and I still struggle with that. You will go back and forth and sometimes or redo things from scratch because you figure out halfway that in order to do X you should have done Y a while ago, and now it's too late. Then repeat this process ad nauseam and finally you have your model... if you're still somewhat new to this... like me 😄

However, it turns out modeling is not the hard part... Currently, I regard texturing to be far, far more difficult. Again you have to learn specialized (sometimes expensive) tools, but that's not the main problem. Good texturing requires a certain sense of artistry, and some people have it and others don't. The likes at FlyTampa and FlightBeam have it. Me? I'm not there yet, suffice it to say. This is, incidentally (I think), why sometimes you see poorly modeled objects that look great, and highly detailed models that look terrible. It's the texturing!

But, even with all that, now try to do all this using the rather unforgiving tools of P3D. In the beginning it did feel like the tools were working against me at every step. With little understanding of how the system works and a SDK documentation that, as @Wolkenschreck mentioned, can be downright vague, sometimes seems conflicting, and in general just isn't very friendly to total beginners on top of old-looking tools that can be hard to navigate... Well.

I will add that the people over at FSDeveloper are excellent. They really do try to help every step of the way. And I'm not saying the tools are bad per se. They are actually excellent - very powerful. ModelConverterX is, honestly, pretty much a one-stop-shop for model-based addon development. Combined with ADE and SBuilder, that's petty much the suite of specialized P3D tools that I personally use (and then there is Blender for modeling, and Substance and Photoshop for texturing). However, it's not easy for beginners to get to grips with them.

Additionally, as has been mentioned repeatedly, you constantly have to reload the sim every time you make a small change. Even if P3D loads pretty fast for me, I probably spend 5-10 minutes every time I have to reload the testing scenario just to see if my small texture change made things better or worse. And this is certainly an important thing to do, simply because the previews you see in your tools don't necessarily translate to the same thing in P3D.

The funny things is that, to some extent, things can all come down to perception. What would you be more inclined to develop for: the very feature-complete but often cumbersome development framework and toolkit of P3D that requires constant reloading to see the effects, or the nicely integrated on-the-fly presentation of MSFS, despite being still rather feature-poor?

In my view, MSFS' SDK is indeed feature-poor, but a bunch of important things can be done just fine. The things that cannot be done as easily are often relegated to default, such as the custom groundpoly (though you see the more experienced development houses, such as FlightBeam and FlyTampa, release MSFS sceneries with custom ground polies already). The thing is: the MSFS defaults look much better than the P3D defaults. And so, with probably half the effort that you make a scenery for P3D as a beginner, you make one for MSFS that graphically outshines the same scenery in P3D.

EDIT: just to add to Rob's point about the MSFS SDK documentation being vague: true! Indeed, MSFS's SDK documentation is a bit of a drama in its current state. And it means that you can't really do anything fancy in MSFS, be it scenery or aircraft. But I do believe that despite that, simply because of MSFS' superior-looking rendering, with little tools a beginner can still make something that's essentially quite nice-looking without as much as hassle as would be the case in P3D.

[Benjamin J 751]

13 minutes ago, dave2013 said:

I agree.  The airport background poly should also conform to the underlying terrain, as well as runways, taxiways, aprons, etc.

There should also be an automated mesh smoothing incorporated into the process, although  this may not be necessary for lower mesh resolutions.  It could be an option.

I recently made an airport for Calama, Chile and it looks awful because of the considerable E-W slope of the terrain - one side either sits atop a very high cliff which looks horrible, or I lower the cliff and end up with a terrain wall at the other end.

Hopefully LM will work on this as it is very important for airport realism.

Dave 

I think that the mesh does indeed conform to the AFCAD made in ADE, if I'm not mistaken. The issue is that he custom ground poly made in Blender or elsewhere, does not. As Rob mentioned, the best thing you can do now is import the mesh into your 3D modeling software and go back and forth between the modeling software, ADE and the sim to match everything up. I have not done it, but the pure idea of doing this was enough for me to not even want to try it anymore.

Though I think it's important to mention one thing: we have had sloped runways in P3D/FS for quite a while. LatinVFR's SLSU was one, for exmaple, many years back. The main difference is that these airports never supported AI, which rely on the AFCAD for the altitude of runways and taxiways. The user aircraft, however, as long as the 3D groundpoly is defined as a 'platform', can make use of it just fine, though perhaps with some issues for the plane's altimeter which I believe does rly on the mesh, and not a custom 3D-modeled terrain...

The funny thing is that LatinVFR's KFLL shipped with a module that allowed AI to recognize 3D models as platforms (I guess, I never looked at how it actually works), and therefor use sloped runways just fine without all the pain of the current system. Which is just to say that there are additions that can be made to make the system easier to work with.

[chapstick 2,757]

Even if you take the time to do sloped runways in P3D, the results are usually not great due to limitations of the engine. For one thing, there are issues with floating objects like taxi signs/trees and see-through terrain. And it can make taxiing totally wonky with aircraft bouncing violently at grade changes and playing their "crashing touchdown" sound. 

[Benjamin J 751]

21 minutes ago, Chapstick said:

Even if you take the time to do sloped runways in P3D, the results are usually not great due to limitations of the engine. For one thing, there are issues with floating objects like taxi signs/trees and see-through terrain. And it can make taxiing totally wonky with aircraft bouncing violently at grade changes and playing their "crashing touchdown" sound. 

True that. Though I think taxi signs etc can be given their own altitude values as well, though i might be mistaken... But yeah, it occurred to me now that Dave's 'smoothing' suggesting might have hinted towards this particular issue indeed, where more abrupt grad changes cause rather violent reactions from the aircraft.

 

@Rob_Ainscough Sounds like something that I will never be able to in my 'career' as a wannabe P3D developer 😅

[Christopher Low 5,339]

20 hours ago, Beardyman said:

I run P3D5 on 1650/6GB of Ram, my vram usage usually is 2.7 to 3 GB with EA on and all my addons - no problem at all, and sim look way better than P3D4

I am curious. In what way does P3Dv5 look "way better" than P3Dv4?

[GSalden 4,656]

28 minutes ago, Christopher Low said:

I am curious. In what way does P3Dv5 look "way better" than P3Dv4?

If you have V5 you can see it when turning on EA...

This you will never get with V4 :

 

But if you are satisfied with V4 then that’s all that counts 😃

[Christopher Low 5,339]

Dusk and dawn effects in v4 already look pretty good on my PC with REX 4 Texture Direct Enhanced Edition installed. If "way better" actually means the atmospheric effects, then I am indeed satisfied with v4