I will reveal a little secret: my password consists only of numbers. Does the keyboard layout affect passwords from numbers? And why did everything work in versions 0.1.7.0 and 0.1.7.5, and now the Windows developer was to blame?
Linux is certainly great, but not everyone can use it and not everyone wants to use it.
And the prevalence of Linux leaves much to be desired.
There is no Windows developer. I am the only core developer working on this entire project, currently. Anyone can contribute code on Git but very few do so. We are always open to working with new developers, too.
It is not a keyboard layout that the issue concerns. Rather, its the way your text is
rendered. Windows uses Wide-set UTF-16 (which is no longer backwards compatible with ASCII). So the fact that it is composed of numbers holds zero difference, as it is still a string of text, not bytes. Just because you read a number doesn't mean that a machine reads it at all.
Our code runs on UTF-8 so that means that anywhere that Windows has to process text, necessitates a special condition and conversion into windows-compatible text. Your machine reading text differently than you do, and needing to convert it, is the reason your password is showing as incorrect.
For those who use computers for devops or anything in compsci really, Windows is nothing more than a huge headache with restraints. Due to this, Windows developers are harder to come by. Windows is intentionally unwieldy. Microsoft's doing. (My opinion, although I'm certain many would agree)