Random832 writes:
There are plenty of standard encodings that do have actual representations of the control characters.
My complaint was not about coded character sets that don't conform to ISO 2022's conventions about control vs. graphic blocks, especially in the C1 block. It was about promoting *unassigned* codes to the Unicode scalars with the same integer values. These codes don't correspond to characters. They are undefined as far as codecs are concerned. In the case of windows-125x charsets, even though they are IANA registered, Microsoft reserves the right to change and even ignore the published repertoire without updating it. There I think it's reasonable to use WHAT-WG graphic character repertoires even in Python's stdlib codecs, and I wouldn't be surprised if Microsoft was willing to delegate definition of those repertoires to the WG in the end.