Thanks for the explanation of why there's no explanation.
I dunno... I've seen those ??? any number of times in the past, but I never thought enough of them to post.
I believe, though I haven't dug through the source code enough to confirm, that there is actually a text encoding bug in UBB. It seems that the code which generates lists of threads in UBB doesn't respect the current setting for character encoding, but the code in UBB that is responsible for displaying posts does.
The short and not-too-technical root of the problem is that there is an accepted international standard way of assigning all the "regular" characters of the English language and all the standard punctuation marks you see printed on a keyboard; for example, a space is always represented by the number 32, the uppercase letter P is always represented by the number 80, and so on.
But when it comes to accented characters, foreign-language characters, special punctuation like smart (curly) quotes, special symbols like the trademark or registered trademark symbol, and things like that, there's no accepted standard way to represent them. There are actually a bunch of different, competing standards: ISO Latin, MacRoman, UTF-8, UTF-16, and so forth.
Naturally, they're entirely different. In ISO Latin, the registered trademark symbol is assigned the number 174. In UTF-8, it's represented by the number 49838 (UTF uses two-byte numbers for special characters).
Normally, the browser and the Web page communicate with each other to tell one another which standard they're using. You can tell UBB which standard to use, and then UBB will tell the browser which standard it's using. If you change the standard you're using, UBB will change stuff around accordingly. But if something goes wrong, which it seems to when UBB tells the browser it's using one standard and then sends post titles using a different standard, the browser will display the wrong character, or the special ? symbol to indicate it can't figure out what that character is.
