Anyway, I skimmed through the adaptive block size limit algorithm

Question

Anyway, I skimmed through the adaptive block size limit algorithm

CHIP and I really like the approach of the algorithm with the control function, elastic buffer and forget factor in particular. I haven't dove deep into the exact maths though. I also like that the limits are not based on some hard coded values or expected growth curve and are purely demand driven 👍

I currently have two questions:
1. What effort would be required to manually change the upper limit if the demand pushes it to levels that are not supported by the network in terms of hardware? Is it easy enough if such a situation occurs?
2. What is the implication on the fees if there are seasonal/abnormal bursts of transactions for sustained periods of time that would exceed the upper bound limit? Like it can be seen in this picture:
@bitcoincashautist

#bch #bitcoincash #cryptocurrency #english

0

20.08.2023

4 ответов

14 просмотров

Arthur Автор вопроса

bitcoincashautist
1. minimal effort, the algo already implements a "...

Thanks. So setting a fixed upper limit if desired would require a change in the software of the nodes? Couldn't it also be changed by miners as a parameter? And which option of the two is more reliable to be activated if needed? 2. That's what I thought. So the curves of the control block size and block size limit would be a bit steeper than in the picture, right? And the fees would temporarily go higher as well until the pool clears up?

0

20.08.2023

bitcoincashautist

Arthur
Thanks. So setting a fixed upper limit if desired ...

it could be set as configurable parameter, but these are not the kind of params you can just tweak without coordination with others, what miners can easily do is just self-limit their own blocks so they don't move the algo beyond that, no need for a hard upper limit

0

20.08.2023

Arthur Автор вопроса

bitcoincashautist
it could be set as configurable parameter, but the...

Would they succeed if they self limit to the maximum block size considered safe or would they need to set a limit considerably lower?

0

21.08.2023

bitcoincashautist · Accepted Answer

1. minimal effort, the algo already implements a "hard" limit but it's 9 exabytes (for implementation safety, so we can't overflow uint64), wouldn't be too much trouble to change that or implement some fixed schedule curve, I had done it already to test that idea 2. in reality that picture is not possible because there the sizes go over the limit, I presented it like that because I wanted to see original data without it being clipped by the simulated limit. What would actually happen is that those bytes would spill over to the next block and make the algo go faster until mempool is cleared

209 похожих чатов

Anyway, I skimmed through the adaptive block size limit algorithm

4 ответов

Похожие вопросы