Progressive encoding and decoding of 'repeated' protobuffer fields

I recently came up with a novel arbitrary-integer-length binary encoding called BLIP (binary large integer prefix), which also begat a novel binary container format (which of course can be of arbitrarily large size)

https://github.com/pmarreck/BLIP

Anyway, a protobuf-relevant excerpt from the README:

    "BLIP is 4x faster than LEB128/Protobuf for large value 
    encoding (1.3 vs 5.1 ns/op) because it writes raw LE 
    bytes with a single memcpy instead of shifting and 
    masking 7 bits at a time."

Among other (I think) interesting features, there's a reserved bit (but only if it makes sense, if the value is multibyte) for endianness, which would potentially solve that whole problem (which is also documented in the README)

I am not really advertising it so much as wondering if it's actually useful to others as I don't normally operate at this level (I'm an Elixir dev by trade, although I seem to be drifting towards much lower levels... Zig)

I recently toyed with a tiny VM for protobuf, you can use it in redpanda WASM streams, or in streaming pipeline (without having to use SAX parsers or IDL shenanigans with projected message compiled...) for filtering and projecting before deserializing. 100% written by claude.

https://github.com/fsaintjacques/wql