IRC channel logs

<rlb>I just hit cases that made me wonder if it might make sense to have some way to (format ...) a bytevector as a bytestring, for example, in error messages for invalid user provided port encodings, or error messages for decoding errors.

<rlb>At the moment, a failed ascii decoding would report #vu8(110 111 112 101 32 182) but I'd argue that since in most cases, the bytes will be an ascii superset, it'd be nicer to put #u8"nope \xb7" in the error message. I'd certainly rather read that.

<rlb>oh, actually it's be #u8"nope \xb7;"

<rlb>If we can add format characters, one option might be to have ~b format bytevectors as bytestrings. Then we can use that in relevant scm_misc_error () messages, etc.

<rlb>"it'd be"

<rlb>I suppose if the noncharacters proposal settles, and we were to decide to include it, could solve the problem another way, i.e. could just convert the bytevector to a string with noncharacters and render that, but that'd be more expensive, and a bit further removed from the original data...

IRC channel logs

2026-09-13.log