I’ve been watching him since I think 2020. He made some of the best tech content of YouTube and I’ve never seen another individual have more knowledge of Windows internals than him. He was accused by Youtube in the past many times for spreading viruses in the description of some of his videos.
This time it was a copyright violation of some random Japanese channel, with no relation to his channel. It is very possible this error was made by the new AI YouTube copyright detection bots.
Please help spread this. RIP Enderman 2016-2025



By the way, there was a video by Enderman (or FlyTech, or similar) showing a Windows locale that looked ﻉกƚٱɼєℓץ ʟ𝔦к𝚎 Շዘノร, intended for English-speaking devs to test support for Unicode and unusually short/long strings in the UI. I haven’t been able to find it for years (the title, which was along the lines of “The Strangest Windows 8 Build” didn’t help). Has anyone seen it recently?
The actual term for it is “Pseudolocalization”, might throw up a hit
Wow, that’s a very informative article! I only knew about Faux Cyrillic, Greek, Vietnamese etc., which are parts of the text transformation, but not about the technique as a whole. I guess I’ll edit the Faux Cyrillic and “faux German” (Metal umlaut) articles to help anyone search for info about a cursed string they saw in a niche setting deep in Windows.
See, I’ve been wanting to make a post about Windows 11 suddenly being like
and I needed the video to provide more context. I guess the Wikipedia article could be enough but it obviously doesn’t show screenshots.
Overall, localization on Windows has gotten worse, there are context blunders that wouldn’t have happened in XP days.
They’ve released at least one screenshot, but since they mostly use it for internal testing, and they very rarely ever release those builds, there’s not much to go on.
So builds they didn’t intend to release provide the best insight.
Now I’ve only ever used the English releases, but I know before Vista that this simply wasn’t a thing. Each language release was a separate build, handled separately by different teams. So there was never a mixing of languages or loading strings at runtime, they were hardcoded into the binaries that shipped on disk.
Sounds like something FlyTech would show.
But I don’t remember that happening.