I'm reporting a handling/conditions anomaly that has persisted since May 17, 2026 on the GTSweatlords server, affecting multiple drivers simultaneously. I've done extensive client-side elimination (including a full Windows reinstall and multiple CSP/Pure versions with and without DXGI) and worked through structured telemetry analysis daily. I've reached the limit of what I can diagnose client-side and believe the root cause is in the CSP/Pure weather-conditions layer.
Summary
On May 17 (Nürburgring 24h qualifying day), connected drivers reported a sudden change in on-track conditions, after which the server went down. Since then a subset of drivers experiences degraded car behaviour that no client-side action has resolved. Critically, client-side telemetry shows the car's own physics values reading normal-to-healthy — which points away from car/setup and toward a shared CSP/Pure conditions state.
Unexpected weather not matching the configured session profile
Different grass height on verges / run-off (GrassFX)
Loss of mechanical grip / traction (as felt)
Reduced high-speed stability and cornering load (reported as "downforce loss")
Universal symptom across ALL users: telemetry can no longer be recorded cleanly (see §3)
Affected setup
Server: GTSweatlords — discord.gg/2PyNRuSYHf — 49.13.117.156:9601 (DE) / Kunos 51.89.20.52:8313
Title: Assetto Corsa + Content Manager + Pure (CSP)
Car/track: Porsche 911 RSR 2017 / Nordschleife Touristenfahrten
Perception split: only a minority of drivers report it (mix of wheel and controller users); most report no change.
Universal symptom — telemetry logging corruption
Across the whole community, telemetry can no longer be recorded normally. It is either:
not recorded at all,
only a brief sliver captured after leaving the pits, or
chopped into fragments with flat/blank sections in between.
This is consistent across users and is, in my view, a strong shared signature of a CSP-side state problem rather than per-user config.
My own measured deficit
Same car, same setup, matched track temp:
Online: ~6 s/lap slower than my pre-incident pace.
Offline: ~20 s/lap slower, with tyres overheating from Turn 1.
Pre-incident reference: a clean 6:07 on May 13 (four days before).
Tests & remediation performed
System / install (all completed, failure persisted):
Steam Verify integrity of game files on Assetto Corsa.
Full machine restarts; network reset.
Full clean reinstall from scratch: Assetto Corsa, Content Manager, and Pure/CSP — all car/track mods removed, reconfigured from a clean state with no migrated configs.
Multiple CSP/Pure versions tested, with and without DXGI.
Full Windows reinstall.
MOZA wheelbase software reinstalled; wheel/pedals recalibrated multiple times.
Re-verified map.ini, sectors.ini, and car data.acd.
Telemetry diagnosis (162-channel MoTeC + live logging), pre- vs post-incident, same Juria setup, matched 14°C road:
Grip is NOT lost — it measures higher: peak lateral G 2.60 g → 2.78 g; median cornering 1.24 g → 1.34 g.
Steering rack torque per g is unchanged (31.4 → 30.3 Nm/g, −3%, within noise) — the physics FFB input is intact.
Surface Grip channel reads a constant 100.000 in every lap, before and after — i.e. the anomaly is not captured by AC's native grip telemetry.
Tyres run ~8–9°C colder than before at the same road temp, yet grip is present.
My data.acd is byte-identical (verified by MD5) to another, unaffected driver's working car, and I run the same CSP version as them.
In-sim FPS is healthy (147–160) — not a performance/frame issue.
Community-level: we have collectively excluded both a purely local error and a simple server error, given the simultaneous, multi-driver onset tied to a weather change.
I'm reporting a handling/conditions anomaly that has persisted since May 17, 2026 on the GTSweatlords server, affecting multiple drivers simultaneously. I've done extensive client-side elimination (including a full Windows reinstall and multiple CSP/Pure versions with and without DXGI) and worked through structured telemetry analysis daily. I've reached the limit of what I can diagnose client-side and believe the root cause is in the CSP/Pure weather-conditions layer.
Summary
On May 17 (Nürburgring 24h qualifying day), connected drivers reported a sudden change in on-track conditions, after which the server went down. Since then a subset of drivers experiences degraded car behaviour that no client-side action has resolved. Critically, client-side telemetry shows the car's own physics values reading normal-to-healthy — which points away from car/setup and toward a shared CSP/Pure conditions state.
Unexpected weather not matching the configured session profile
Different grass height on verges / run-off (GrassFX)
Loss of mechanical grip / traction (as felt)
Reduced high-speed stability and cornering load (reported as "downforce loss")
Universal symptom across ALL users: telemetry can no longer be recorded cleanly (see §3)
Affected setup
Server: GTSweatlords — discord.gg/2PyNRuSYHf — 49.13.117.156:9601 (DE) / Kunos 51.89.20.52:8313
Title: Assetto Corsa + Content Manager + Pure (CSP)
Car/track: Porsche 911 RSR 2017 / Nordschleife Touristenfahrten
Perception split: only a minority of drivers report it (mix of wheel and controller users); most report no change.
Universal symptom — telemetry logging corruption
Across the whole community, telemetry can no longer be recorded normally. It is either:
not recorded at all,
only a brief sliver captured after leaving the pits, or
chopped into fragments with flat/blank sections in between.
This is consistent across users and is, in my view, a strong shared signature of a CSP-side state problem rather than per-user config.
My own measured deficit
Same car, same setup, matched track temp:
Online: ~6 s/lap slower than my pre-incident pace.
Offline: ~20 s/lap slower, with tyres overheating from Turn 1.
Pre-incident reference: a clean 6:07 on May 13 (four days before).
Tests & remediation performed
System / install (all completed, failure persisted):
Steam Verify integrity of game files on Assetto Corsa.
Full machine restarts; network reset.
Full clean reinstall from scratch: Assetto Corsa, Content Manager, and Pure/CSP — all car/track mods removed, reconfigured from a clean state with no migrated configs.
Multiple CSP/Pure versions tested, with and without DXGI.
Full Windows reinstall.
MOZA wheelbase software reinstalled; wheel/pedals recalibrated multiple times.
Re-verified map.ini, sectors.ini, and car data.acd.
Telemetry diagnosis (162-channel MoTeC + live logging), pre- vs post-incident, same Juria setup, matched 14°C road:
Grip is NOT lost — it measures higher: peak lateral G 2.60 g → 2.78 g; median cornering 1.24 g → 1.34 g.
Steering rack torque per g is unchanged (31.4 → 30.3 Nm/g, −3%, within noise) — the physics FFB input is intact.
Surface Grip channel reads a constant 100.000 in every lap, before and after — i.e. the anomaly is not captured by AC's native grip telemetry.
Tyres run ~8–9°C colder than before at the same road temp, yet grip is present.
My data.acd is byte-identical (verified by MD5) to another, unaffected driver's working car, and I run the same CSP version as them.
In-sim FPS is healthy (147–160) — not a performance/frame issue.
Community-level: we have collectively excluded both a purely local error and a simple server error, given the simultaneous, multi-driver onset tied to a weather change.