String ordering difference between Windows 10 and version 1803

Hi all,

We've noticed some differences in string orderings between windows 10 enterprise version 10.0.16299 and version 1803, specifically affecting accented characters. Both OS versions had the same system and input locale (en-gb) and time zone (UTC+00:00). Has anyone else encountered this and know whether this is intended or not?

Repro steps:
* In powershell, create a text file including some accented characters - we used the upper and lowercase a-grave characters:
> "a" > .\characterFile.txt
> "A" >> .\characterFile.txt
> "a'" >> .\characterFile.txt
> "A'" >> .\characterFile.txt
> "a-" >> .\characterFile.txt
> "ab" >> .\characterFile.txt
> "a-b" >> .\characterFile.txt
> [char]0x00C0 >> .\characterFile.txt
> [char]0x00E0 >> .\characterFile.txt
* Still in powershell, run "Get-Content .\characterFile.txt | Sort-Object". We saw differing string ordering depending on the OS as follows:

In Windows 10 Enterprise:
A
a
À
à
a'
A'
a-
ab
a-b

In Windows 1803 version:
A
a
A'
a'
a-
À
à
ab
a-b

Specifically, the differences are in the orders of a' and A', as well as the location of the accented characters.

Is anyone else seeing this behaviour or know whether this is intended? Any information is appreciated!

Thanks,

Asha