Hello @roadapathy, for all practical purposes, a continuous rebase of Proton onto upstream wine is not a practical use of Proton dev time. All dev time would be consumed trying to keep up with problems arising from subtle changes introduced by upstream wine commits instead of working on bringing improvements to the point releases of Proton.
If you can isolate specific commits that improve compatibility with a game, then cherry picking those commits into Proton can be considered on a per-case basis.
Oh okay, that makes sense — I didn’t factor all of that in. I’m not a developer, so I misunderstood how the Bleeding Edge branch actually works. I assumed it was more like a test playground for Proton (and sometimes it doesn’t even build for me when I don’t touch anything!).
You explained this really well.
I did frame this as a feature request, but my real motivation was that Arc Raiders ran extremely poorly — around 20–30 FPS — when games that are far more demanding graphically (like Star Citizen) run at 120 FPS on the same system. I later confirmed that Arc Raiders runs much smoother on a newer Wine 10.17 build, so I wanted to raise the possibility that Proton might just be missing some of the performance improvements introduced in that range.
That said, I now see this probably belongs more under a bug report than a feature request.
From what I can tell, the biggest relevant upstream changes between 10.03 → 10.18 were:
• NTSYNC – expanded kernel-level synchronization replacing esync/fsync, reducing thread contention in high-thread workloads.
• Scheduler / server optimizations – reduced epoll overhead and improved pipe I/O performance.
• WOW64 / syscall routing – fewer transitions between 32-bit and 64-bit layers.
• win32u / kernel32 improvements – better handling of WaitForMultipleObjects, critical sections, and inter-thread signaling.
Anyway, thanks for taking the time to explain all that — I appreciate the insight, and I’ll go ahead and skedaddle! I'd send you a free coffee for this but I didn't find any contact info for you. lol
Nothing extracted yet.
Feature Request
I confirm:
contain this feature already.
Description
Could Bleeding Edge be considered for the latest versions of Wine? From 10.0x to 10.18, there have been significant changes in performance, support, and features as well as bug fixes. NTSYNC, Wow64, OpenGL, Wayland, X11, etc.
Justification [optional]
I presume Bleeding Edge was meant to be a trail blazer but this might be a difficult change. Wine 11 is just around the corner so this would lay the tracks for that big moment.
Risks [optional]
I tried to drop into the Proton project just the new Wine 10.17. Nope. It broke everything but I'm also not a developer.
References [optional]
I have been building Wine for a non-steam game (Star Citizen) for a year or two now and I see changes in performance and compatibility in the game from each update. I found it significant enough to bring this to the table here at the awesome Proton team.
Thank you!