Having some familiarity with Vietnamese, I found this historical perspective on an old DOS text editor for the Vietnamese language interesting. It’s an impressive piece of software, and highlights some issues with internationalization that persist even today.
asthasr, you beat me to it. I was going to submit it with this comment from jimmie on Hacker News. It’s one of few times I thought about just submitting the HN comment itself.
I like this from that thread:
They still have to… even with Unicode, fonts are often “shaky” in their support of Vietnamese characters, so, for example, Đường Đồng Khởi may render in some fonts with the tone marks as separate characters altogether. The dominant form of Vietnamese input on Windows is a third-party program (Unikey), and the input system on Linux likes to crash every few minutes when it’s being used…
It’s introduced as a text editor, but if it’s handling and displaying more than 256 characters in a single file, it can’t store them as plain text files, so it must have used some special format. Wouldn’t it have been obvious that something weird was going on when other tools couldn’t read files that BKED had produced?
Actually it just displays more than 256 characters and in the context of the article that means graphical chars for the UI. Now that it renders it all on its own, there’s no real need for a special format. I don’t know about the formula and graph examples shown at the end. Those probably had their own format.
But the way it has been written, it can store plain text files. That’s one of the main points of the article IIUC