Arrow11Crash

@Flappie it doesn’t generate a zip, mate - DCS crashes immediately without reporting an exception or error in the log, the log file itself is truncated. I’ll see if I can find one of my truncated ones from earlier today and post it here.

Yes, happens to me as well. Was last working well on 14 Feb 24 and there was a PTC update on the 15th (from the point of view of an Australian time zone). Reverting from PTC fixes the problem. Same issue, the log stops at the point where it lists the options. After a deep clean/repair of DCS, removal of mods, deletion of fxo/metashaders, I initially "fixed" this problem by changing the name of my DCS folder in Saved Games to something else (on the advice of Bigity on the dcs-help-room thread on ED's discord, thanks mate), and in adding things back in one at a time I found that options.lua was the source of the problem. On editing settings one at a time it was apparent that enabling the "Use VR headset" option results in a crash to desktop around the end of when the 15th anniversary DCS splash screen disappears. It will probably need to be raised on the Oculus forums, but it's notoriously hard to make something stick on those forums (you just get mindless AI bot responses from Meta) so we should probably all report it there.

Any chance for a reply from the dev team please?

The full EDM for the F1 has ~600k triangles, but unfortunately both the level 1 and level 2 LODs have identical triangle counts and the level 3 LOD which should be *drastically* smaller still has something like 200k triangles for an object visualized 10 km away. I assume this is a mistake as it's a terrible use of resources. Compare this to the F-14, for example which is a primary model of similar complexity (700k triangles), reducing to 300k, 68k then 1k only for the distant LOD. The LOD distances are a bit different for the F-14 but still, at 1500m an F-14 is 68k versus 600k for the F1.

On my system the performance drops when looking at cities are still present after today’s patch. I can’t tell if it’s worse or the same. It’s such a bloody shame. 3080 Ti, 12700k, 64gb, Quest 2, openXr with toolkit and FFR on. I test this with the F1 cold and dark instant action. Turn openXR toolkit FPS monitor on, load, look right (away from city), look left (towards city). Mine changes from 15-18 ms up to 28-30ms looking at the city - I’ve been unable to set up anything that causes such a bad performance drop anywhere else, even Marianas with the Apache is better than that.

Man it would be nice if this was the cause of the micro stuttering many of us get. Thanks for putting in the time to collect data and put it here.

Update: this is related to the ‘VR settings not saved’ bug reported elsewhere by another user. If I go into the VR tab at the beginning of a DCS session and I click/reclick the appropriate checkboxes the controllers no longer grab the stick and throttle. Given that this is apparently happening for the other VR options it looks like DCS is doing something odd with the options in terms of the active game state.

My hand controllers snap-grasp the stick and throttle when they are nearby, if my palm grip button is pressed. I’m on a Q2, set with hand controllers active and requiring a grip press, but with all the other checkboxes off (e.g. not to use the stick or throttle with the hand controllers). Latest OB. I haven’t been messing with core DCS files or any mods to controls. Have repaired already. I don’t recall this being an issue a few versions ago but can’t point to an exact version where it started. It’s a real problem particularly in the hornet when manipulating the centre MFD. It will even grasp a hidden stick. I have enquired on discord and this behaviour seems aberrant but I have no idea why my setup would be different. I can also trigger cockpit mirrors with my head the mirage and the ejection overhead lever in the F1 with my head. Is this a known bug? Is there a workaround?

This is still happening and it's very frustrating. It happens with head back during dogfights etc, looking over shoulders. It's clearly firing the top ejection handles based on head position. I wonder if Aerges is handling the Oculus controllers/head position in an unusual way compared to other modules because there are other irritating problems (e.g. using the hand controller thumb stick to rotate dials is difficult at times depending on the actual orientation of the controller relative to the control, which *should not happen* - it should depend on the headset position and orientation within the cockpit).