This is by design, so that case conversion and folding is just a bit operation.
The idea that SOH/1 is "Ctrl-A" or ESC/27 is "Ctrl-[" is not part of ASCII; that idea comes from they way terminals provided access to the control characters, by a Ctrl key that just masked out a few bits.
I guess it's an age thing, but I thought this was really basic CS knowledge. But I can see why this may be much less relevant nowadays.
It’s on the list:
<https://web.archive.org/web/20251103035213/https://www.catb....>
I've been in IT for decades but never knew that ctrl was (as easy as) masking some bits.
You can go back maybe 2 decades without this being very relevant, but not 3 given the low level scope that was expected in CS and EE back then.
I learned about from 6502 machine language programming, from some example that did a simple bit manipulation to switch lower case to upper case. From that it became obvious that ASCII is divided into four banks of 32.
Been an ASCII-naut since the 80's, so .. its always amusing to see people type 'man ascii' for the first time, gaze upon its beauty, and wonder at its relevance, even still today ...
Yes, the diagram just shows the ASCII table for the old teletype 6-bit code (and 5-bit code before), with the two most significant bits spread over 4 columns to show the extension that happened while going 5→6→7 bits. It makes obvious what was very simple bit operations on very limited hardware 70–100 years ago.
(I assume everybody knows that on mechanical typewriters and teletypes the "shift" key physically shifted the caret position upwards, so that a different glyph would be printed when hit by a typebar.)