Author: Karl Johan Alm 2017-06-01 02:11:48
Published on: 2017-06-01T02:11:48+00:00
The writer of the message has been researching block filter digests, with a view to optimizing bandwidth, space and resource usage. The report on this research can be found at http://bc-2.jp/bfd-profile.pdf. The author provides a graph showing the results of simulations which demonstrate how bandwidth is used by different wallets when synchronizing once per block, once every 144 blocks or across all wallets. One interesting finding was that if digests are allowed to contain multiple blocks, the false positive rate of high block count digests can be higher than normal. The author seeks comments on their approach, assumptions and possible math mistakes. They also ask about acceptable costs in terms of bandwidth use and memory, querying whether field sizes restricted to 128 MB are appropriate for short-lived fields, or whether 256 MB fields might be better for over-filled, high false positive rate fields. The author also wonders whether serialization is overkill or not.
Updated on: 2023-06-12T01:29:26.676949+00:00