%oJ9A++vrbxXlCgGLqttaVh5eFmQBA/rqpkBiq5QXHkY=.sha256Dev Diary 19/10/2018
- started drafting a post on merkle-trees for ssb logs, but got stuck on the actual implementation details, so I postponed it because I also want to get some programming done today: https://gist.github.com/AljoschaMeyer/a824a410032fcf8ad310cf80c24a185e
- and prior to drafting, this involved a bunch of reading and thinking
- I think we can do better than this structure: https://github.com/blockstack/blockstack-core/issues/146
- this adds quasilinear space overhead to the feed (I think?), a tree-based solution should manage to get it down to linear
- this is just plain amazing: http://soc1024.ece.illinois.edu/gpads/gpads-full.pdf
- seriously, I did not expect that searching the web for
merkle skip listwould lead into such an amazing rabbit hole- which I'll have to explore in my free time I guess - back to programming for now
- but just imagine combinding λ• with Lasp... there's just so much CS still has to explore =)
- seriously, I did not expect that searching the web for
- I think we can do better than this structure: https://github.com/blockstack/blockstack-core/issues/146
- and prior to drafting, this involved a bunch of reading and thinking
- finished the serde json serializer
- does not yet have a mode for automatically serializing map entries in the correct order though
- nearly finished a serde json deserializer
- more merkle trees - no point in fighting my own brain
- I'm starting to accept the need for a merkle skiplist rather than a pure tree, but I'm still unhappy with the space requirement in the feed itself...