bech32-optimized QR codes #12191

@sipa said in #11991:

I haven't looked at the code, but inside QR codes it's advantageous to use uppercase Bech32 (it will result in a smaller QR encoding).

Is that still the case when rendered as URI?

Should amounts be added in a different way?

Would that involve amending BIP-21?

Bech32 addresses are valid both in all-lowercase or all-uppercase, so even inside a URI you can use the uppercase version (though the bitcoin: locator should remain lowercase).

But wasn't there a more efficient but not BIP-21 compliant way to encode amounts?

No clue. All you need to get the better efficiency in QR is sequences of several characters of uppercase+digits in a row. It doesn't require everything to be uppercase.

@sipa

No clue. All you need to get the better efficiency in QR is sequences of several characters of uppercase+digits in a row. It doesn't require everything to be uppercase.

~That is not the case. For better efficiency with alphanumeric character encoding all symbols must belongs to the set 0..9, A..Z, %, *, +, -, ., /, :.~

Ref: https://en.wikipedia.org/wiki/QR_code#Encoding

@hebasto My understanding is that QR codes encode a sequence of groups, where each group has a single encoding, and a variable number of characters. As long as there are sufficient numbers of characters to emit a separate group, libqrencode will use it.

@hebasto From the wikipedia article:

Encoding modes can be mixed as needed within a QR symbol. (e.g., a url with a long string of alphanumeric characters )

@sipa TIL, thank you.