Why is it that emails written in non-Latin languages sometimes display correctly, while other times they just appear as squares and question marks?
Given the nature of the Internet, your email might potentially be handled by software which is decades old. Protocols for handling email are very old, and that means that your email might potentially be handled by software that can’t deal with any form of Unicode at all—or any other non-ASCII character set for that matter. The content of email can be mangled by such software. Newer protocols have been designed to avoid this kind of problem in handling character sets, but because of the widely distributed nature of the email infrastructure, potential points of failure of character set conversion may exist for decades yet, until everything is using Unicode correctly. There is no way to diagnose a particular email problem without exact details of scenarios, because so many pieces of software and so many different protocols are involved. These kinds of problems can creep in at any of the interfaces between them—or sometimes even internally to some particular piece of software. Simply sayin
