[mcguire]

Thanks for this. For those of us not following closely, that paper explains the issue clearly.

From my reading, the protocol is supposed to adjust the difficulty of the block chain proof of work to provide a certain number of blocks per hour?

[riprowan]

Yes, that is correct.

[gravypod]

From what I understand from reading many of your comments (riprowan, seems like you understand what you are talking about):

  - Some people don't want to increase the block size
  - The original whitepaper says that this is supposed to increase
  - People who advocate for this have some big players trying to hinder their discussion

I don't know if I am getting this right. But does that sound about what is happening here?

[riprowan]

point 1 is obviously true

To point 2 - the original white paper doesn't even mention "block size limit" because originally Bitcoin did not have a block size limit. "Honest miners" are assumed by the author to always include all profitable transactions (ie. where the fee > marginal cost to include it in a block).

The limit was added after an attack in the earliest days of Bitcoin. That attack vector is simple: a miner who mines a block includes bogus transactions to himself in order to inflate the block to gigantic proportions. Other miners, who have to validate its contents, choke on the block. Meanwhile the attacker is free to keep mining, and / or the network stops processing transactions.

The original attack, and why it's no longer relevant, is described in this blog:

http://gavinandresen.ninja/One-Dollar-Lulz

In short, the attack vector which prompted the creation of a block-size-limit consensus rule is no longer a concern, as it has been addressed in two ways:

(1) This attack surfaced in the earliest days of Bitcoin, when it cost nothing to mine and miners risked nothing if they mined a block and threw it away by attacking the network. Today, industrial miners risk $10K with every block they would lose if they tried the original attack - which by the way

(2) is preventable without even having a consensus rule, for example by running Bitcoin Unlimited, a version of Bitcoin which attempts to re-create "original" Bitcoin by allowing each miner to set his own "anti-attack" limit, instead of requiring the network to agree to share the same limit.

It is clear from the various online records that the limit was intended a non-economic anti-attack threshold, to be removed if it ever became an issue.

[gravypod]

I wish you wrote a blog post about this, I'd like to see what perspective you would bring to the table.

A lot of this makes little sense to someone from the "outside" like me, but you are able to explain it in a good way.

Thanks for the help!