At one time Lua supported Unicode characters into identifiers, especially vowels with accents like it's common in Italian and Spanish. Also "c" with cedilla which is used in many Turkish words and other languages. But eventually that support had to be removed because it was causing a lot of trouble. Not necessarily because people used the interpreter on computers only capable of running WindowsXP or even older. In many cases there was C/C++ code to maintain and that required fighting with "locale" setting, which was usually set to English. That sucked for many people whose first language isn't English, which forced them to use their preferred language only in comments and which also forced them to use text editors that supported Unicode.
Julia is truly a modern programming language but one has to remain in that camp entirely. For example, using VSCode/VSCodium to compose programs for it. Many symbols used in Julia were actually demanded by mathematicians and scientists and people like that; they would have refused to use a programming language that didn't otherwise, no matter eg. how many places and how fast it could compute the circle-number. QB64PE and Lua rather recall the past. Python might be somewhere in between. It might not support Unicode in identifiers because many programmers are lazy, they just want the code to work and their great nightmare is hours of debugging.
I could reveal my experience with a music-creation application that I had used a lot 10 years ago but sadly, cannot use it any longer because it "loves" WindowsXP. Actually there's an "alpha" release of it now but the developers changed a lot of stuff, and have to use it on clunky Windows10 and therefore I'm not comfortable with it anymore. This application developed a problem showing the open-file requester because of an Unicode problem. It took about three years and three minor releases of the program for the developers to get to the bottom of it. This problem was a hinderance to work flow in that application. The head developer's favorite language is Spanish (maybe one of the dialects from Spain) which might have had something to do with it, as he had to deal with the C/C++ "locale" set to English as was necessary because the developers were stuck using Microsoft Visual Studio.
Julia is truly a modern programming language but one has to remain in that camp entirely. For example, using VSCode/VSCodium to compose programs for it. Many symbols used in Julia were actually demanded by mathematicians and scientists and people like that; they would have refused to use a programming language that didn't otherwise, no matter eg. how many places and how fast it could compute the circle-number. QB64PE and Lua rather recall the past. Python might be somewhere in between. It might not support Unicode in identifiers because many programmers are lazy, they just want the code to work and their great nightmare is hours of debugging.
I could reveal my experience with a music-creation application that I had used a lot 10 years ago but sadly, cannot use it any longer because it "loves" WindowsXP. Actually there's an "alpha" release of it now but the developers changed a lot of stuff, and have to use it on clunky Windows10 and therefore I'm not comfortable with it anymore. This application developed a problem showing the open-file requester because of an Unicode problem. It took about three years and three minor releases of the program for the developers to get to the bottom of it. This problem was a hinderance to work flow in that application. The head developer's favorite language is Spanish (maybe one of the dialects from Spain) which might have had something to do with it, as he had to deal with the C/C++ "locale" set to English as was necessary because the developers were stuck using Microsoft Visual Studio.