If your input is limited to ASCII, sure.

But ASCII is only a 7-bit standard, and only supports those characters needed by American English computer users in the 1960s. Lots of characters you might see in “plain text” are not part of ASCII; including all accented characters, all non-Latin alphabets, and many common symbols and punctuation marks including these: £€¢©™°

(Yes, you could get accented characters in the pre-Unicode days using 8-bit “extended ASCII”, e.g. IBM/Windows code pages. However, those are not really ASCII and they will break if the text is interpreted as the wrong code page.)

Unicode collation is the Right Thing today.