[spuz]

Very interesting report. I'm not sure I understand what the root cause of the issue is though - is it a Linux bug or a problem with the monitor itself?

[pwg]

The monitor reports an EDID resolution/refresh value to the system that is incorrect for the actual monitor's capabilities. Linux happens to be choosing to use that invalid EDID resolution/refresh value and the default result is no picture on the monitor.

The bug is with the EDID values LG programmed into the monitor.

[bombcar]

There are many cases where hardware is not up to spec and other operating systems ignore it more or less but Linux can be a stickler.

Windows is generally most forgiving, then Mac, finally Linux.

[wtallis]

More like: Linux is most likely to try to use the features a device claims to support. Windows will often only try to use a narrower subset of those features, and that subset is what actually got tested before the product shipped. Case in point: NVMe APST, which has been a perennial source of trouble on Linux, but Windows largely ignores (at the expense of worse power management behavior).

[jldugger]

I sort of expect that Windows and Mac have more testing and overrides applied to fix buggy firmware. The find and override process just happens during pre-ship QA instead of post-launch support issues debugged over the internet.