mSparks

they did and simheaven did. https://simheaven.com/x-world-3-0/ This year, Microsoft added estimated heights to the global buildings footprints, for Northern America, Australia and Europe (as June 2023), maybe more to come in future. In total there are now 1,200 millions footprints and 174 millions estimated heights available

It would be no different than what sim heaven is doing now. In fact simheavens latest Xworld already includes a good chunk of it, because it was released under a permissive licence. The problem is the same now as then, overwealmingly the quality is bad and there's no night lighting, hence asobo launching an army of designers into "world updates". see also

maybe, not convinced, the only thing "new" here Im seeing is the ability to add some code that does more than "is this bit of data in the field of view" before the gpu decides to do calculations on it. Not sure I trust how any of the theoretical explanations translate into actual code. And even that is just "standardising" it and dedicating some silicon to it, because devs have been doing that in the compute shaders since there were compute shaders. I had a quick look around to see where its up to, been a while, seems Alan Wake 2 beat Laminar to being the first to do it themselves: https://www.pcgamer.com/mesh-shaders-explained/ reading that, I'm not even sure you can call them mesh shaders on anything not running DX12U. Which is backwards and somewhat comedy, because nanite on console is PS5 exclusive.

pretty much exactly what I said at the start. ->fastest is mesh shaders, which is one of several very sexy technologies that become (more widely) possible as they finish up getting the fundamental tech working they discussed in the dev q&as I posted earlier, while nothing has been officially announced here, I suspect/expect they will. Austin has talked around it a few times

I had something similar happen to me when I refreshed all my custom scenery and set up serving it all from zip files. A blanket install of xworld hits perf pretty hard. If you are fps chasing there are two or three of the "high detail" xworld folders you want to disable, and there is some procedure for integrating with XPs new forests properly. The documentation on this is good though from what I remember, Ill try and remember to start a thread on it when I get back next year. Still possible its something else related (e.g. changes they made to the 3D veg since the initial release), they very recently fixed up shadows off the 3D veg, and that fix definately had a bit of a perf hit, maybe its worse at very high screen volume.

think that is one of the xworld scenery packs, there is some stuff about that in the readme iirc.

for XP there is debugging for the water mesh activatable from art drefs, you can turn on the wireframe, that looks like mesh shaders to me. pretty sure the new trees are to. aiui mesh shaders are where the majority of the mesh "only" exists in the GPU and is created by a shader. And yeah, I also get there is more to it than just this. no way they are sending all that geometry over the bus like they do with the conventional scenery. As for metal, Austin has a fairly tight relationship with Apple, he was one of the few people who knew about the first iPhone before it was announced, which is how he got xplane on it almost for launch day. (aiui)

I'll put aside the sarcasm because it is a valid question. mostly because by default there is a lot of numerical geometry in the ground textures, 10s of thousands of technically 3D surfaces each with their own texture. gpu + cpu APIs are pretty slow at this, and havent improved it for some 20 years. the .fac files they use for many of the buildings arent that much better. faster is ortho - that replaces it with a well defined single object with LOD and one texture. MS does this by default. fastest is mesh shaders, which is one of several very sexy technologies that become (more widely) possible as they finish up getting the fundamental tech working they discussed in the dev q&as I posted earlier, while nothing has been officially announced here, I suspect/expect they will. Austin has talked around it a few times, this will probably be part of MSFS2024. Aces of Thunder aleady demod it hiding in the background of their announce trailer pretty sure XP12 is already using it (or very similar) for the new water stuff that went live with XP12.

scenery also has a habbit of being one of the common culprits. especially ancient plugins loaded from the scenery folder. other here is 3X monitors at 4K is going to need FSR dialing up to handle the insane number of pixels getting thrown around. wouldnt surprise me is 2020 is turning DLSS on automatically.

it works fine in 2d as well, I just wrote it for VR, that hit works on closest match, e.g. in that 744 script looking to the left of the mcp, my two rotaries do ias and heading, if looking to the right of the mcp they do Alt and VS, if looking down it does com1. I use those $2 ardunino rotaries, so they have push button functions as well. It took a bit of tweeking to get comfortable in the R44, but I use similar there for qnh, radio heading com and nav radios (which is what I wrote it for originally), the math was fun, size of spheres tangential to a line, closest match is the smallest sphere. which simplifies down to almost nothing, couple of cos sins and a square root.

If you get a minute, would be very interested to know how its perf is stacking up against the sparky744.... Because when I started the sparky744 zibo mod wasnt a VR friendly aircraft, but it's had a lot of development since then.

for a really simple version, lua supports assigning functions to a variable. e.g. function iasup() -- increase IAS end local upFunc = iasup upFunc:once() that script I linked does it by finding the function assigned to the closest x,y,z point in the direction being looked at. if you are using buttons you can just do it directly from thier function. if you struggle Im back at my desk on like the 5th of Jan or smth, it'll only take 5 minutes or so to make you a lua script you can drop in any aircraft that uses xlua. if you want a look version all you need is the x,y,z of the knobs you want to turn and drop it in the hotspot array. (I have 2 rotaries, hence two functions)

I dont know what if any plans they have for low end machines, there is XP11 for them until they upgrade. Well, there are lots of factors that affect performance, most 3rd party stuff never cared (I got at least 3 bans from the org over the years arguing with their mods that 25fps isnt enough - before pulling my finger out and doing something about it myself), the last gen XP default scenery that ships for now isnt perf friendly at all. Sparky744 + ortho + autoatc is pure joy in VR. As are all the VSkyLabs products own. They covered a fair bit of their plans there in a previous dev q&a going back a few months. e.g. not sure if the others are public. transcript of the second one is tho https://docs.google.com/document/u/0/d/13JkolLoebjceCGurDspg3ioijLhUuX-V9gKVDvcJGb0/mobilebasic#heading=h.lb4ymtifd380 They've been working closely with several commercial setups on complex visuals, sure bet any issues there are a top priority. No idea where ms is up to, Ive looked at their dev tools a couple of times and they basically seem completely unusable garbage to me (and not just because its all windows only). Not that it matters now, 2020 dev tools are DOA and 2024 dev tools arent public. I guess that whole transparent communication thing everyone was raving about being so much better than Laminar stopped working out for them. Dont need that in my life, my sympathies to anyone stuck in that hell, even if there is a hint of told you so on my part. I will be quite surprised if msfs 2024 does not hit massive delays tho. especially after aces of thunder on psvr2 starts pawning the flight sim game market

https://www.x-plane.com/2023/10/coming-in-x-plane-12-0-8-and-beyond-roadmap-update/ dwindling devs isnt even close to true, there are more devs working directly with LR now than was claimed working for Asobo at the height of MS's development. XP isnt P3D that was clearly 100% dependant on a single 3PD (PMDG). Which highlights another misnomer I often hear. (successful) 3PDs dont "follow the money". They follow the best development tools that allow them to create the highest quality products. Then the money follows the highest quality products.

if you mean what I posted, its a linux backtrace. Its a "standard" linux tool, that among other things reports exactly where in code it is stopped, in that case the function inet_shutdown() waiting for a thread to finish. which although that is written by LR, Im guessing is short for internet shutdown. There is very likely more than one cause, AutoATC had plenty of them over its development - afaik AutoATC related hanging are all resolved now, the really hard thing with stuff that happens "only sometimes" is understanding the root cause can take ages.

Like I said, xplane is in kinda a weird place atm. No ones buying (or starting) XP11 projects. XP12 isnt quite there in terms of number of people who own it and the developer tools to kick off big new projects. Well, Tony would be better placed to confirm or deny the scenery thing, but good quality scenery doesnt look like any less work than a good aircraft to me, architect's can spend years designing one building and its surroundings, and a typical airport has many structures and vehicles that need recreating. Plus sales of scenery are going to be much lower than an aircraft even in optimum circumstances, which doesn't exist anywhere (other than maybe DCS). I can't imagine there is any more active development specific for 2020 than there is for XP11 right now. Both are EOL with only a short remaining lifespan, that doesnt mean ongoing projects suddenly get shelved, look how long FSX kept limping along. It does mean in 2 or 3 years there wont be new products being released for them (because they weren't started today), afaik no one outside of the microsoft partners are making 2024 stuff.

There is no short term It takes a few years from starting even a fairly simple flight sim project to delivering it. just look how long it took pmdg to drop their first aircraft, and they were working on them a full year or more before MS first released. And thats for projects with a full time commitment not ones being done "on the side". What is coming to market now is stuff started mid 2022 at the latest, many people even now still believe xbox is a bigger market than PC, blindly accept it is true because microsoft rather than grasping that its a fraction of the size of PC with sales of xbox consoles falling at 52% year on year. It got so bad last year MS had to put most 3PDs on salary and NDAs to keep them around. The sim bringing users to 3PDs is unsustainable, 3PDs need to bring users to the sim. MS and Asobo know this well, its why they put them on salary. Conversely, vskylabs, hotstart have clearly been knocking it out the ballpark. I expect similar is true for the likes of tolis, flight factor xcrafts and most of the other fully independant XP devs It's very easy to neglect how long a lead time these things have, The Sparky744 went byccnc4 in May 2019 - "just" starting to get traction, and that was for an aircraft that already had heavy dev for XP10 and XP11

there is no special xbox exclusive forum for msfs, they use the same as PC: https://forums.flightsimulator.com/t/navigraph-flightsim-community-2023-survey-is-out/617200 The truth is, xbox series X/S that can actually use it have not sold well (a few million units), only 3 million in 2023 by some estimates https://www.ign.com/articles/ps5-has-outsold-xbox-series-x-and-s-3-to-1-in-2023-new-data-reveals VR is the future. Wouldnt be surprised if aces of thunder on psvr2 tops the 2024 survey. msfs for the other "100s of millions" of xboxes gets something like a 1/10 rating for play anywhere. And even for the new series X|S is somewhat mixed opinions https://forums.flightsimulator.com/t/does-msfs-run-well-for-xbox-users-top-model-now-here-in-late-2023/614924 very similar to xplane mobile. - lots and lots of users tried it, many even bought some content, but it is not something they take seriously enough to visit any related websites, let alone complete a survey. You can see this just from the fraction of users that write stuff, entirely consitent with the results of the survey. This is absolutely why imho MS are going for round 2 with 2020 + 4 + missions. They realised very very late in the day they need something to make xbox users come back rather than uninstalling it after the first session (drunkedly flying over their house) . Do you have any evidence that contradicts it? because I don't. The msfs2020 "shock cancelled" and msfs2024 announced in between dec 2022 and dec 2023 is also entirely consistent with the total xbox flop seen in the 2022 results, and the disastrous xbox console sales in the last 12 months.

the sparky744 includes MCP hotspots written in lua that change a command up/down function based on what you are looking at. https://github.com/mSparks43/747-400/blob/0173f76ce4eba7f14175972610cc08dc6955e5e3/plugins/xtlua/init/scripts/B747.99.VRcontrols/B747.99.VRcontrols.lua#L269 should be fairly easy to repurpose that to your needs.

thats a very narrow question tho, that could probably be reworded to "which flight sim did you read about the most in the last 12 months" and get the same answer. for the bigger picture you need to see what people were doing when they were not firing up X flight simulator because Y developer announced yet another fix for their broken payware. that tells a very different story about how things have changed in the last 12 months. Dec 2022 Dec 2023 And that is all before XP12.1.0 drops https://www.x-plane.com/2023/10/coming-in-x-plane-12-0-8-and-beyond-roadmap-update/ Which I (am fairly sure I) said from day 1 will be the first version "most" start using as their primary sim. I note all the commentary on how many millions of flight sim "pilots" there are has also now vanished like a trump in the wind, I'd love to know how those who believed in that figure reconcile it with the results of this survey, even at 13% that 10 million figure would still be 1.3 million XP12 full price sales in the last 12 months, there is no $1 option for x-plane.... plus "flight simmers that have XP12" is now a smidgen under 40%. Anyone offering any insight into whether they think ms2024 manages a better delta in the next survey? Because I think thats gonna be crunch time.

vast majority of this stuff is still free atm, Ill invest in some marketing when there is some chance of a return. Until then I'm just taking feedback on what works and what doesn't. that video worked better than most

aside from needing a LOT more high quality .mil airframes and scenarios. This, imho, is really where the "x-plane approach"/planemaker carved itself out a solid future. Not relying on flight test data but instead getting rock solid estimates of how it would fly based solely on the design & documentation (and a ton of validation and tweaking against actual flight test data) opens the door to successfully building stuff like https://www.vskylabs.com/p/vskylabs-test-pilot-sr-71-tb-project/ We are a long way yet from the CL650/PMDG/zibo/sparky standard being the standard. We are even quite a way off https://forums.x-plane.org/index.php?/forums/topic/176949-kmgm-montgomery-regional-airport-payware-quality-freeware/ being a base standard. I think we see much more of that next year and those following.

it was smth like 800,000 3 years ago. the vast majority are either not real people (bot accounts), the same person multiple times, or just registered an account to download & try something with the demo. Note their most ever online was way back in 2020. The numbers Goran posted are far more realistic, about 5000 a year a good result (thats still $500,000 to divide up for a $100 CL650..), at 5% of the market that would be 100,000 as a high end estimate. As another data point, latest sparky744 release has had roughly 20,000 downloads since the start of August, + 5000 downloads of the XP11 backport. The total size of XP has grown nicely from last year (my estimate for last year was about 200k), as new people started XP12, and only 5/35 XP11 users switched to XP12, carried on shrinking a fair bit for MS, which is now less than twice the number of XPlane unique users. but we are talking 10s of thousands of people, not millions.