[Bitcoin-ml] What i would like to see in the next HF + timeline

[Chris Pacia]

Regarding restructuring the merkle tree for parallel validation.. It
just occurred to me that if this allows transactions to be stored out of
order on disk then it will likely break Bip37 as the the
BLOOM_UPDATE_ALL flag requires dependencies to be processed first so
that the child will then match the filter.

I think we really need to put our heads together and come up with a
better replacement for Bip37 anyway, but this is at least one
consideration.