X-Plane 11.35 beta 4 is now live – run our updater and check “get betas”. Release notes here.
Steam users: we’ll release this tomorrow if there isn’t a sign of a huge fire overnight – it’s already uploaded to the servers.
Add-on Developers: if you haven’t run your add-on in 11.35, please put it through its paces now. Everything that might be risky or weird is already in and we’re just killing off bugs now. If your add-on has a compatibility problem, we need to know now if we’re going to fix the problem during beta. While we’ve tried to make everything “just work” with old add-ons, there’s always the chance that your add-on does something special we didn’t think of. Please test now!
43 comments on “Add-on Developers: Please Try X-Plane 11.35 Beta 4”
Is this update available for non steam owners of X-Plane 11?
It’s available via the LR installer for LR customers now – we’ll put it on the Steam beta today if we haven’t gotten any “my GPU caught fire” bug reports overnight. Since Steam users who opt into the beta are auto-updated (rather than clicking “get updates” when prompted) we try to give the beta a day to make sure we’re not going to brick people’s installs.
So when will the update be available for the rest of X-Plane 11 owners?
It’s live now – has been for a few days.
Could be a 11.35b4 issue ?
The model that i have made has a normal map for reflectivity , when i use this model in under XP windows the reflections are correct on the windows , but when i use it under XP linux the whole building has a reflection. Thought this is an error on my model , but even the freely released 500D has that same reflection on everything in linux only.
I had filed a bug couple of weeks ago where i was noticing that even the smaller buildings are showing reflections.
I can recheck it tonight if it was just a one time thing or it’s really a bug.
It sounds like a driver bug. You have to check a few things:
1. Are the rendering settings the same? is one on minimal?
2. If you run 11.34 on the SAME install (install without beta) does the artifact go away or persist?
3. Are the numeric driver versions the same between Linux and Windows.
If all of those things are true, please file a bug.
Will reupdate the drivers for linux and check.
Both XP (windows / linux) use the same settings.
If the issue still persists a bug will be filed.
Yep think it’s a driver issue , the reflections now seem to have gone.
I am trying to change the preset position in VR but this did not work as expected.
BEGIN_TELEPORT_HOTSPOT SITTING Pilot’s Seat
AABB -0.55 -0.65 -0.90 0.0 0.67 0.3
PRESET_XYZ -0.2438 0.42 0.0
When I used the reset the pilots position on the VR controller I thought I would be looking at the left wing but instead I was still looking straight ahead. I am going on what I found here //developer.x-plane.com/article/aircraft-vr-configuration-_vrconfig-txt-file-format-specification/
I don’t know whether it helps, but did you try “PRESET_PSI -90.0”?
In further testing today I found out this.
If you use the “Get to Pilot’s Seat” on the touch controller it will ignore the PRESET_PSI setting.
On the other hand if you teleport to the pilot seat location it will use the PRESET_PSI setting.
So my question is this a bug or expected behavior?
Will wait for Ben to chime in here.
This sounds intentional, like, you can’t fly the aircraft facing backward. 😉
Ah. But what if it’s a Sikorsky Skycrane? Just shows that there’s at least one use case for this.
We need the ability to face backwards – not only with the VR config, but with the Plane Maker positioning fields as well.
Should I file a bug report/feature request, Ben, so that this can be tracked. Seems like it would be an easy bit to code. I’ve actually made the psi adjustment manually using a text editor in the ACF, and X-Plane loads it just fine. Unfortunately, every time I go into Misc Objects in Plane Maker, it undoes my edit. 🙁
Thanks for considering the change!
You can file it but we aren’t going to get to it in 11.35.
Getting it in 11.35 would be sudden service that I wouldn’t expect. Just pleased that you’re willing to consider inputs like this at all, Ben – it’s great to have the level of interaction that we do with the team. Many thanks!
Hi Ben, Bill and Steve,
Interesting discussion. It would perhaps make sense to differentiate the “move to pilot’s seat” command and the “Reset VR View” command. The “Move To Pilot’s Seat” would do the rotations (PSI=PRESET_PSI), while the “Reset VR” would not (PSI =0).
It is probably important to keep the “Reset VR View” as it is now, to be able to reset all the axis in the case where the coordinate system in VR gets messed up after the user does some outside teleportation…
I believe that this discussion arose because of the fact that for some users the projected image of the controllers in VR sometimes don’t seem to align with the view. Or for some people their view is consistently rotated when they do a reset. In that case they want to apply a default rotation of some degrees to be perfectly aligned with their cockpit (another instance where it is useful to have a PRESET_PSI !=0).
Crash after New Mission start in Linux still present (XPD-10173). According to gdb, stack trace indicates crash is still in fseeko, just as my bug report detailed earlier (complete with included gdb log showing the stack trace). Because this is a crash, I can perform no further testing under Linux until this is fixed. Reverting to working Beta 2.
Unrelated to this beta, but I was wondering for some time about the increasing size of the airports file(s) (>300MB already):
It seems X-Plane loads them all into memory as text file (or does it parse them all and keep the parsed data)? Anyway since X-Plane is using memory-mapped files (I guess it still does), I could imagine that the airport files are basically kept on disk, while once per update of the airport file(s) binary indexes are built for:
Airport by global (intergalactic 😉 location
Airport by ID
Airport by name
These indexes would refer to the position of the corresponding airport in the text file.
On start of X-Plane those indexes would be read or memory-mapped themselves, while the airport data file(s) are just memory-mapped.
When needing airport data, the program would fetch the position from one of the indexes and then “read” them with that offset from the mapped airport file, causing “page in” of the corresponding file blocks.
That should preserve some memory and reduce starup time significantly.
Maybe there are even more clever algorithms…
We are already doing this – have been since at least 10.0!
OK, so something else must be quite slow when loading. Maybe at least the Log.txt could contain some detailed timings for loading the parts of the simulator…
Another thing I frequently see with Windows (7) is that the splash screen appears, then loading starts, the screen becomes black until loading has finished. Especially if loading takes quite long that doesn’t look nice.
I notice this too. Often times if I click on the the screen while X-Plane is loading, Windows wants to give me an application not responding error. I imagine that something is blocking the UI thread.
this is do to the sim not reporting back to windows that its alive blame its mac roots
X-Plane not pumping the Windows event queue has nothing to do with the Mac.
It’s been some time since the last time I tested X-Plane 11, now in VR – really cool, but I struggle with the performance.
Btw: The new FMOD sounds for the King Air are great!
Just some quick questions…
i) Do the Baron 58 still use the old sound system?
ii) The ground is still jumping around a little bit, I thought this had been fixed in an earlier build?
iii) Clouds are really really bad for the performance, is there perhaps a way to optimise the system in later versions? Why are clouds so much of a performance problem?
iv) Any ideas or insights in regards to performance optimisations/modernisations in general you could share with us?
B58 is FMOD sound. Maybe I can do a separate post on tuning performance.
Would be great with an update on performance. BTW: When running X-plane in VR, would it be of any use to add an extra GPU, one for each eye?
One GPU per eye is *not* supported for x-Plane.
Ok – btw.: it was anti-aliasing that killed the frame rate with clouds, at the lowest anti-aliasing-setting, cloud rendering was no problem
Actually several generations ago, clouds started to kill the frame rate, especially in X-Plane 10 (independent of VR). I had to reduce cloud complexity significantly for a reasonable frame rate. What makes things worse in X-Plane 11 is that you cannot control the cloud settings via GUI any more (besides of having to guess what actually kills you frame rate). Especially “almost overcast” (still a little transparent) seems to have a big effect.
Maybe my question is a little bit off topic,
asking on X-Plane forum I didn’t get an usefull answer.
What is the dataref “sim/operation/create_snap_marker” good for?
Is it still in use with (in, at?) the current version?
I didn’t find any info how to use this dataref.
It’s not a dataref, it’s a command that puts down a marker in the replay that you can see on the vertical map.
Note sure what code has changed recently but performance has really gone South with latest beta’s. I’ve tried just about everything, clearing the preferences, removing add-ons, installing older Nvidia drivers, but in dense locations the fps plummets (more than usual) and there’s stutters.
X-Life version 3 plugin was the framerate killer. It usually has minimal impact but with beta 4 it kills fps by 50%.
Excellent work from you and all of the team in these releases! Have watched the recent seminars from FlightSimExpo online and its great to see the constant progress and updates being worked on.
Slightly off topic but with the release of the new 3rd Generation of Ryzen CPUs (eg 3700x, 3900x), im curious if the team has done any testing with these new CPUs and what kind of frame performance is being achieved in comparison to existing Intel CPUs. Hopefully the increasing core / thread count combined with increased IPC will translate to better performance down the line, but I’m considering a system upgrade to new hardware so improved parity to intel would be nice!
For the time being, X-Plane’s performance is bottlenecked on a single thread. (See past discussions on Vulkan for more about the future of this stuff.) While the new Ryzen CPUs look awesome, you shouldn’t expect much in the way of performance gains for X-Plane compared to, say, an existing high-end i7 or i9.
good news is the new 3×00 line from AMD has single thread IPC = or better then gen 9 intel finally
Ryzen/ Xp11 benchmarks! //www.comptoir-hardware.com/articles/cpu-mobo-ram/39273-test-amd-zen-2-x570-a-ryzen-7-3700x-ryzen-9-3900x.html?start=14
new AMD stuff has as good or better IPC then gen 9 intel so they area good buy now even wile stuck in single thread
Do you plan to introduce other antialiasing mode to X-Plane? I am mostly interested in Temporal AA.
Not any time soon, but I agree temporal AA could help e.g. in the HDR render.
It could help enormously. Really hope you guys one day decide to implement it.
Is this beta ever going to be final?
(Although there has been internal discussion about simply never finishing the beta so we wouldn’t have to fix the beta bugs. There has also been internal discussion about never _starting_ the beta, so we wouldn’t get the beta bugs in the first place.)
Comments are closed.