Re: [bitcoin-dev] Draft BIP : fixed-schedule block size increase

[Adam Back <adam@cypherspace.org>]

Note Jeff's proposal and Greg Maxwell's flexcap proposals also have a
growth limiter as a hard-cap and a mechanism for influencing a dynamic
cap within that envelope.

The hard-cap serves the purpose of a safety limit in case our
understanding about the economics, incentives or game-theory is wrong
worst case.

More comments to follow.

Adam


On 25 June 2015 at 15:50, Gareth Williams <gacrux@gmail.com> wrote:
> On 24 June 2015 1:49:51 PM AEST, Jeff Garzik <jgarzik@gmail.com> wrote:
>>Miners can collude today to lower the block size limit.
>
> Of course they can. What, then, is the need for BIP100's hard-limit voting mechanism?
>
> You only need consensus rules to enforce block size limits if you're enforcing them _against_ miners. Which may be a perfectly valid thing to do (if your threat model includes, for example, the possibility that large miners deliberately create large blocks to gain an advantage over small miners.) But BIP100 doesn't address that anyway.
>
> Wouldn't it be safer for consensus to get behind Gavin's simpler 8MB->8GB hard-limit growth curve*, and then encourage miners to enforce a soft limit below that, agreed through a voting mechanism? The later can be implemented at any time without consensus changes -- nobody can prevent miners from coordinating the max block size they'll build on anyway.
>
> * but with a safer "supermajority" than 75% please :)
> --
> Sent from my Android device with K-9 Mail. Please excuse my brevity.
> _______________________________________________
> bitcoin-dev mailing list
> bitcoin-dev@lists.linuxfoundation.org
> https://lists.linuxfoundation.org/mailman/listinfo/bitcoin-dev