I don't think this is a good way to go about solving the mentioned issues and view it as a step backwards. I would much rather see an actual fix to the mentioned problems instead of reverting per-monitor DPI awareness. IMO the user experience without per-monitor DPI awareness is worse than a window placement bug.

This is a blocking issue and want to resolve quickly for folks to move forward. I think there are other issues with the actual fix of per-monitor dpi aware and needs more investigation. If we don't want to revert and continue investigation on higher priority, i'm good with it. If we can't get a fix out by next release, we should consider reverting the change to unblock people to move to latest electron. @kevinsawicki your thoughts?

This is a blocking issue and want to resolve quickly for folks to move forward.

I dare to doubt the gravity of the bug as something that would "block" a way "forward" for people. Can you be more specific? Note that per-monitor DPI awareness has been turned on for a couple of releases now.

Is there any update or progress on this issue?

I think we should close it. This does not provide a fix to the aforementioned issues.

@alespergl So, what are other solutions? I would love to see this fixed as I'm currently experiencing this scaling issue (Win 10, 125%).

This was put up for review in one of our standups. We would normally discuss this in out monday meeting, but most of us are in Tokyo next week, so perhaps an in-person conversation can be had there, if not we can make a final decision at the next standup.

VS Code is badly broken on Windows due to all the window position issues. Would like to see this merged eventually 👍

@bpasero This PR is not a fix for your problems, it's a workaround. VS Code can have this workaround applied internally - pm me if you need help with that.
We need to close this PR and replace it with a real fix without regressing high DPI funcionality.

@alespergl yes, thanks for clarifying. I would prefer the real fix too if possible :-).

This pull request has been hanging for too long, and apparently we don't have an agreement here. I'm closing this pull request and let's continue the discussion at #10862.