[SuperNinKenDo]

>we introduced a change in Qubes 4.2.0 that caused inter-qube file-copy/move actions to reject filenames containing, e.g., non-Latin characters and certain symbols

Crikey. I'm sorry but what will it actually take to get the notion that "ASCII is enough" out of Western programmer's heads? I've been interested in Qubes for a while, but long suspected stuff like my reliance on non-Latin character sets and other issues would make it a non-starter. The only thing surprising is that I'm not surprised.

[fhsketbx]

> out of Western programmer's heads

This is less about regional borders and more about security.

Security is all about compromise.

I use Qubes because it’s an operating system where the compromise always heavily favors security over other considerations. Limiting to ascii shrinks the attack surface.

The fact that ascii contains English characters is historical rather than ideological.

The team behind Qubes is multinational, but the lead is Polish and the Polish language has a character set outside the ascii standard.

How do you propose protecting against the sort of attacks Qubes was trying to mitigate?

[SuperNinKenDo]

I think a whitelist that covers more than English would be the way to go. You could even set up maintainers for each block or something.

I don't think it's ideological in any "conscious" sense, however there is an argument to be made about ideology being the water the fish swims in.

Also I know I said ASCII as a shorthand and because that's often something that developers say, but it's quite possible these "Latin" characters aren't pure ASCII but may include extended Latin sets like Latin-1 and the like, which might cover the Polish case.

Again though, I think that you could have a system where you have maintainers for different blocks to audit them for whitespace or other deceptive characters and that could work fine.