Why can't I write U"\U00000020"?

[martinho....@gmail.com]

I just came across §2.3p2 which concerns universal character names and reads "(...) Additionally, if the hexadecimal value for a universal-character-name outside the c-char-sequence, s-char-sequence, or r-char-sequence of a character or string literal corresponds to a control character (in either of the ranges 0x00–0x1F or 0x7F–0x9F, both inclusive) or to a character in the basic source character set, the program is ill-formed."

This means that I cannot write U"\U00000020", and must instead write U" " or U"\x00000020". I cannot find a reasoning for forbiding the first form. Can anyone who knows a good reason for this chime in?

[Richard Smith]

You appear to have misread the text you quoted. That is valid, because
it is within the s-char-sequence of a string literal.