[bitcoin-dev] A summary list of all concerns related to rising the block size

[bitcoin-dev] A summary list of all concerns related to rising the block size

[Jorge Timón]

We've identified a fundamental disagreement in:

- The block size maximum consensus rule serves to limit mining centralization

But as said I believe at least 2 different formal proofs can be
produced that in fact this is the case. One of them (the one I'm
working on) remains true even after superluminal communication, free
unlimited global bandwidth and science fiction snarks.
But let's just list the concerns first.

I believe there's 2 categories:

1) Increasing the block size may increase centralization.

- Mining centralization will get worse (for example, China's aggregate
hashrate becomes even larger)
   - Government control in a single jurisdiction could enforce
transaction censorship and destroy irreversibility of transactions
      - Some use cases that rely on a decentralized chain (like
trustless options) cannot rely on Bitcoin anymore.
      - Reversible transactions will have proportional fees rather
than flat ones.
         - Some use cases that rely on flat fees (like remittance) may
not be practical in Bitcoin anymore
- The full node count will decrease, leaving less resources to serve SPV nodes.

2) Trying to avoid "hitting the limit" permanently minimizes minimum
fees (currently zero) and fees in general

- If fees' block reward doesn't increase enough, the subsidy block
reward may become insufficient to protect the irreversibility of the
system at some point in time, and the system is attacked and destroyed
at that point in time
- Miners will continue to run noncompetitive block creation policies
(ie accepting free transactions)
- More new Bitcoin businesses may be created based on unsustainable
assumptions and consequently fail.
- "Free transactions bitcoin marketing" may continue and users may get
angry when they discover they have been lied about the sustainability
of that property and the reliability of free transactions.

Please suggest more concerns or new categories if you think they're needed.

Re: [bitcoin-dev] A summary list of all concerns related to rising the block size

[Thomas Zander]

On Wednesday 12. August 2015 12.28.39 Jorge Timón wrote:
> But let's just list the concerns first.

Concerns?

I have never heard of "develop-by-concerns"? Is that similar to 
fire fighting management?
To that I have this reply;
http://www.aleanjourney.com/2009/07/stop-fighting-fires.html


Your question makes sure that the proper answers don't fit.
And that may be why you are getting frustrated because I've given a lot of 
reasons why a blocksize increase is needed soon, and none of them show up 
in your list...

http://lists.linuxfoundation.org/pipermail/bitcoin-dev/2015-August/010120.html

http://lists.linuxfoundation.org/pipermail/bitcoin-dev/2015-August/010129.html

And various others; I'm not going to bother pasting URLs, you probably have 
them already.

Point is, Bitcoin is a growing network, with a growing amount of transactions 
and we KNOW we will get into problems when the block consistently get full. 
How do we know? Because of the old guys in this list having the experience 
that this always happens, because anyone in the IT business can tell you the 
same.

We have started LN to cope with this growth, but this won't be enough since it 
just won't handle all usecases and thus the on-chain growth will continue. For 
instance remittances and cross-border purchases.
We need both LN as well as bigger blocks.

-- 
Thomas Zander

Re: [bitcoin-dev] A summary list of all concerns related to rising the block size

[Milly Bitcoin]

> I have never heard of "develop-by-concerns"? Is that similar to
> fire fighting management?
> To that I have this reply;
> http://www.aleanjourney.com/2009/07/stop-fighting-fires.html

A set of goals is needed and then you develop road maps.  At that point 
you can propose specific changes that fit within a road map.

Russ

Re: [bitcoin-dev] A summary list of all concerns related to rising the block size

[Jorge Timón]

On Wed, Aug 12, 2015 at 6:41 PM, Thomas Zander via bitcoin-dev
<bitcoin-dev@lists.linuxfoundation.org> wrote:
> On Wednesday 12. August 2015 12.28.39 Jorge Timón wrote:
>> But let's just list the concerns first.
>
> Concerns?

Yes, concerns or risks. Feel free to bike-shed the word.

> I have never heard of "develop-by-concerns"? Is that similar to
> fire fighting management?
> To that I have this reply;
> http://www.aleanjourney.com/2009/07/stop-fighting-fires.html

I'm not the manager of Bitcoin Core nor the Bitcoin consensus rules.
I'm just trying to separate "dogs from cats" (if you allow me the
analogy) in another random attempt to move the discussion into more
productive territories.
I wouldn't be doing this if I didn't felt we have advance identifying
and understanding each other's concerns better.
If you think this thread is useless you don't need to participate.

> Your question makes sure that the proper answers don't fit.
> And that may be why you are getting frustrated because I've given a lot of
> reasons why a blocksize increase is needed soon, and none of them show up
> in your list...
>
> http://lists.linuxfoundation.org/pipermail/bitcoin-dev/2015-August/010120.html
>
> http://lists.linuxfoundation.org/pipermail/bitcoin-dev/2015-August/010129.html
>
> And various others; I'm not going to bother pasting URLs, you probably have
> them already.
>
> Point is, Bitcoin is a growing network, with a growing amount of transactions
> and we KNOW we will get into problems when the block consistently get full.
> How do we know? Because of the old guys in this list having the experience
> that this always happens, because anyone in the IT business can tell you the
> same.
>
> We have started LN to cope with this growth, but this won't be enough since it
> just won't handle all usecases and thus the on-chain growth will continue. For
> instance remittances and cross-border purchases.
> We need both LN as well as bigger blocks.

This thread is for "guardian dogs". To discuss to mouses catches we're
missing, I've created the other almost-identically-titled thread:
http://lists.linuxfoundation.org/pipermail/bitcoin-dev/2015-August/010188.html
The title of this one doesn't contain a "not" like the other one. You
missed (imagined?) that little detail.

In other words, if you want to contribute to this thread, think about
"things that could go wrong if we rise the consensus block size
maximum".
Judging from your strong conviction about the need of an urgent size
rise, probably you have thought hard about this before concluding that
there's not risk for your favorite proposed size (2MB, 8MB, 8GB,
20GB?).
Note that even if you are certain that those risks aren't a concern at
this scale, identifying and somehow estimating or measuring (if that's
even possible) the risks will be useful for future increases.