Talk:Digital encoding of APL symbols
Underscored alphabetics:
When APL was first made available on IBM printing terminals in the 1960s, the typing element did not enough room for lower case a-z. Thus underscored alphabetics were an available way to get another "case". Possibly the Unicode designers thought that underscoring was a text attribute, not unlike the underscore, bold, italic, etc. as one would find in MS Word. This is not true with APL - underscored characters were distinct from non-underscored characters and allowed a kind of upper / lower case. Only A-Z and the "delta" symbol could be underscored. With the IBM 3279 display terminals, one could now have three alphabet cases in programs - upper and lower case, plus underscored characters.
Usage of underscored alphabetic characters are considered by many APL programmers to be obsolete, not modern, and in bad style. APL+Win has eliminated them from their version of the language. Other implementations still support them, but consider them deprecated.