Has anyone thought of implementing a new data structure for

Question

Bitcoin Cash (BCH)

Hossein

Has anyone thought of implementing a new data structure for

mempool to enable parallelization, (To run verify unspent outpoints & tx validation in parallel)?

#bch #bitcoincash #cryptocurrency #english

0

25.07.2024

8 ответов

19 просмотров

bitcoincashautist

tl121x
This needs to be done. This is the key remaining ...

Стикер

0

4 месяца назад

Hossein Автор вопроса

tl121x
This needs to be done. This is the key remaining ...

I was thinking about what is needed to have parallel execution of transactions from the time it is received until it's added to a block. My thinking is that we have to remove the need for chained transactions in a single block. This feature adds extra (unnecessary) rules in tx verification & block creation. The solution I'm thinking involves having a new type of transaction that wraps vector of transactions in it where one can spend outputs of prior txs. And having the ability to replace these types of transactions only if they add new transactions to them.

0

4 месяца назад

molecular#123 🐓

Hossein
I was thinking about what is needed to have parall...

Maybe talk to @fpelliccioni , @mtrycz or any of the guys doing https://scalingbitcoin.cash/en

0

4 месяца назад

minisatoshi

Hossein
I was thinking about what is needed to have parall...

Should also share here https://t.me/bchbuilders

0

4 месяца назад

Hossein Автор вопроса

minisatoshi
Should also share here https://t.me/bchbuilders

There is a discourse website for bch developers. Initially intended to write a post there, But I wasn't able find the website.

0

4 месяца назад

minisatoshi

Hossein
There is a discourse website for bch developers. I...

bitcoincashresearch.org

0

4 месяца назад

Tom

molecular#123 🐓
Maybe talk to @fpelliccioni , @mtrycz or any of t...

to be honest, they are just re-tracing my footsteps of what I did from 2016 to 2020. 🤷‍♂️

0

4 месяца назад

tl121x · Accepted Answer

This needs to be done. This is the key remaining bottleneck that prevents full parallelism so that the full performance of today’s multi- core CPUs and NVMe SSDs can be brought to bear on fast transaction and block processing. Unfortunately, this is a hard problem, because there are many edge cases involving synchronization, such as resynchronizing the mempool at block arrival and orphaning. Eventually, all the low hanging fruit will be picked and this will get done, but only if people don’t spend too much effort adding features and tweaks, especially protocol features that change and complicate the blockchain structure. When this has been done then the critical mining nodes can verify large blocks rapidly, which will enable faster mempool and block propagation. This will increase 0-conf safety and speed and will allow lowering block times while still keeping a low orphan rate. Mining nodes can afford large server configurations with extra performance that keeps orphan rates down, while other nodes can run on slower systems that can be ten times slower and still safely follow the chain.

209 похожих чатов

Has anyone thought of implementing a new data structure for

8 ответов

Похожие вопросы