MIME-Version: 1.0
In-Reply-To: <CADm_WcYwErO1Av_DkMecATQEMFKL7TNZc1Nbs88k-yEKN2vbsQ@mail.gmail.com>
References: <CADm_Wcb+5Xo3HS-FNUYtCapVpYfVvUS_fxpU0Q=TZHJW1=iAFQ@mail.gmail.com>
	<CAAS2fgQOi0amBnPK8Ac3iGDN9CP-mLW6o0ncYdSAOAaqSboejg@mail.gmail.com>
	<CADm_WcYS-zbNFQJ5EPqqkQ5NhgoQNQAgs-SaF_ZZr0QCNFA3_w@mail.gmail.com>
	<CADm_WcYwErO1Av_DkMecATQEMFKL7TNZc1Nbs88k-yEKN2vbsQ@mail.gmail.com>
Date: Thu, 3 Sep 2015 17:57:48 +0000
Message-ID: <CAAS2fgS=sBYscxa+xzi6d2h61HxgeRfdaTj55ospc3oKYgKOXg@mail.gmail.com>
From: Gregory Maxwell <gmaxwell@gmail.com>
To: Jeff Garzik <jgarzik@gmail.com>
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable
Cc: Bitcoin development mailing list <bitcoin-dev@lists.linuxfoundation.org>
Subject: Re: [bitcoin-dev] block size - pay with difficulty
Precedence: list

On Thu, Sep 3, 2015 at 2:40 PM, Jeff Garzik <jgarzik@gmail.com> wrote:
> Expanding on pay-with-diff and volatility (closing comment),
>
> Users and miners will have significant difficulty creating and/or predict=
ing
> a stable block size (and fee environment) with pay-with-diff across the
> months.  The ability of businesses to plan is low.  Chaos and
> unpredictability are bad in general for markets and systems.  Thus the
> binary conclusion of "not get used" or "volatility"

Sorry, I'm still not following.  I agree that predictability is important.

I don't follow where unpredictability is coming from here. Most (all?)
of the difficulty based adjustments had small limits on the difficulty
change that wouldn't have substantially changed the interblock times
relative to orphaning.

> It's written as 'a' and/or 'b'.  If you don't have idle hashpower, then p=
aying with difficulty requires some amount of collusion ('a')
> Any miner paying with a higher difficulty either needs idle hashpower, or=
 self-increase their own difficulty at the possible opportunity cost of los=
ing an entire block's income to another miner who doesn't care about changi=
ng the block size.  The potential loss does not economically compensate for=
 size increase gains in most cases, when you consider the variability of bl=
ocks (they come in bursts and pauses) and the fee income that would be asso=
ciated

What the schemes propose is blocksize that increases fast with
difficulty over a narrow window. The result is that your odds of
producing a block are slightly reduced but the block you produce if
you do is more profitable: but only if there is a good supply of
transactions which pay real fees compariable to the ones you're
already taking.  The same trade-off exists at the moment with respect
to orphaning risk and miners still produce large blocks, producing a
larger block means a greater chance you're not successful (due to
orphaning) but you have a greater utility.  The orphing mediated risk
is fragile and can be traded off for centeralization advantage or by
miners bypassing validation, issues which at least so far we have no
reason to believe exist for size mediated schemes.

As you know, mining is not a race (ignoring edge effects with
orphaning/propagation time). Increasing difficulty does not put you at
an expected return disavantage compared to other miners so long as the
income increases at least proportionally (otherwise pooling with low
diff shares would be an astronomically losing proposition :)!).

Pay-for-size schemes have been successfully used in some altcoins
without the effects you're suggesting.