The whole encoding thing was so annoying. I had experience with this when learning computing, in the 90s in Latvia. The common use case was needing both Latvian

The whole encoding thing was so annoying. I had experience with this when learning computing, in the 90s in Latvia. The common use case was needing both Latvian and Russian languages, at least the ability to read them. Neither was convenient. Russian had its own Cyrillic encodings, but of course there wasn't a single one. Much of the time it was Windows-1251, which you can still run into when browsing the Russian internet even if UTF-8 dominates. But sometimes it would instead be KOI8-R, a different encoding. So you had to know how to convert encodings with special tools, and of course you needed fonts that had the right glyphs in the first place. As an extra fun note, Windows-1251 uses the code point 0xFF for the letter я, which could break poorly-written software that used 0xFF as a sentinel value of some kind. And then on top of that, there was dealing with Latvian. It's a small and relatively obscure language, using an extended Latin alphabet, with Latvian letters like ā, ž, ļ and ķ. So that meant another encoding, Windows-1257, and needing the fonts - some Latvian letters like ž or č existed in many fonts due to also existing in some bigger languages, but ā or ļ were less common. Unlike Russian, Latvian documents would at least be understandable if rendered with the wrong encoding, but input was a bigger problem. Windows had no out-of-the-box support until Windows XP. So from Windows 3.1 up until XP, additional third-party software was required to type Latvian, and the software would do things like hooking into the keyboard driver, which was of course not always side-effect free. There are things that have gotten worse in computing, but the victory of Unicode is a huge step forward. Up until fairly recently, software was very obviously geared towards the 26-letter Latin alphabet, which means English (or Malay).