Return-Path: Received: from smtp1.linuxfoundation.org (smtp1.linux-foundation.org [172.17.192.35]) by mail.linuxfoundation.org (Postfix) with ESMTPS id 8D1E48D7 for ; Tue, 22 Aug 2017 20:06:04 +0000 (UTC) X-Greylist: whitelisted by SQLgrey-1.7.6 Received: from mail-wm0-f53.google.com (mail-wm0-f53.google.com [74.125.82.53]) by smtp1.linuxfoundation.org (Postfix) with ESMTPS id 21A611AC for ; Tue, 22 Aug 2017 20:06:04 +0000 (UTC) Received: by mail-wm0-f53.google.com with SMTP id b189so1760863wmd.0 for ; Tue, 22 Aug 2017 13:06:03 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025; h=mime-version:sender:in-reply-to:references:from:date:message-id :subject:to:cc; bh=8D8AZdnFjMg1mUZNZ1rrmtaRuiHhOFkdqW5HYWrKaAY=; b=bGTKQv0YUsZGnNp54SyDSqUs4RwRUT18EBUpAaRR4ZOQTkR4JibH9MvgJt29pMs4n1 pnUP3bBiY36Q8uLfLnTqzkq/6jyLwDwiuNwlUflaeYgq08q51U8r1Ek1KGzH/KBZ98bl S6td0lyMbwdggNB4T98GVHy37+ZBSQZm0m3XVRni+DlQCuVpC4dGq1eeNmcz7Ed2/8yB /Yl69Vwnfs/toDPoDEpKX/rPWldA/WSN88G8nUec5C4JIfSA/OAOFv2MW3lkPdN/0TJs rvL4syXLLsSPgLkLBvD+rwtpcVMA+qAhy9xtno6k49sQt2uxhOuT7t5U0/CJdcBYTssa M9JA== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:mime-version:sender:in-reply-to:references:from :date:message-id:subject:to:cc; bh=8D8AZdnFjMg1mUZNZ1rrmtaRuiHhOFkdqW5HYWrKaAY=; b=r6Ak0rz3QSxmuTU9zCCiv17l+YsnGddb6PoEx8dUEvGzq0L+jloW6wWiQacdhQu1Wj LPashq9Rig9gpB+Dd145btEEUbJoCJFahdXljqlRAApem91hfdxQ3dOTXg3HNeHZcj2I cyCetXz0GHX81AYv615kmoENGTf5LXiAMvNO997wxdo+dj95IFk8RStDF/5xPoyR1lGq OUH687z32Nh4iqD8qXSJacA8kVhulhupZUd5RiEJXnbYdzISdvfUxADDqPe0sLxnO541 36mov7VcKmrAtBXTFWI+W5VuY6wxaOx9oHmTxY6E0IxLPeI+PGwRsdCTOQhVKVOWLfo3 WVLw== X-Gm-Message-State: AHYfb5jKO29kJn0UdvIUw0kkjN0LCBlJVykoTu1nRqVCdjFhtnXdGZta kHv7hBSRgK84VZQeeHf77ByXssuwUA== X-Received: by 10.28.136.8 with SMTP id k8mr367896wmd.67.1503432362485; Tue, 22 Aug 2017 13:06:02 -0700 (PDT) MIME-Version: 1.0 Sender: earonesty@gmail.com Received: by 10.28.225.135 with HTTP; Tue, 22 Aug 2017 13:06:01 -0700 (PDT) In-Reply-To: References: <4c39bee6-f419-2e36-62a8-d38171b15558@aei.ca> From: Erik Aronesty Date: Tue, 22 Aug 2017 16:06:01 -0400 X-Google-Sender-Auth: 6NMa2yNsWVUZTiIZ2QljehBdBm4 Message-ID: To: Chris Riley Content-Type: multipart/alternative; boundary="001a114434fae8f03605575d1f1c" X-Mailman-Approved-At: Tue, 22 Aug 2017 20:10:48 +0000 Cc: Bitcoin Protocol Discussion , Matthew Beton Subject: Re: [bitcoin-dev] UTXO growth scaling solution proposal X-BeenThere: bitcoin-dev@lists.linuxfoundation.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Bitcoin Protocol Discussion List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 22 Aug 2017 20:06:04 -0000 --001a114434fae8f03605575d1f1c Content-Type: text/plain; charset="UTF-8" > The initial message I replied to stated: Yes, 3 years is silly. But coin expiration and quantum resistance is something I've been thinking about for a while, so I tried to steer the conversation away from stealing old money for no reason ;). Plus I like the idea of making Bitcoin "2000 year proof". - I cannot imagine either SHA256 or any of our existing wallet formats surviving 200 years, if we expect both moores law and quantum computing to be a thing. I would expect the PoW to be rendered obsolete before the Bitcoin addresses. - A PoW change using Keccak and a flexible number of bits can be designed as a "future hard fork". That is: the existing POW can be automatically rendered obsolete... but only in the event that difficulty rises to the level of obsolescence. Then the code for a new algorithm with a flexible number of bits and a difficulty that can scale for thousands of years can then automatically kick in. - A new addresses format and signing protocols that use a flexible number of bits can be introduced. The maximum number of supported bits can be configurable, and trivially changed. These can be made immediately available but completely optional. - The POW difficulty can be used to inform the expiration of any addresses that can be compromised within 5 years assuming this power was somehow used to compromise them. Some mechanism for translating global hashpower to brute force attack power can be researched, and consesrvative estimates made. Right now, it's like "heat death of the universe" amount of time to crack with every machine on the planet. But hey... things change and 2000 years is a long time. This information can be used to inform the expiration and reclamation of old, compromised public addresses. - Planning a hard fork 100 to 1000 years out is a fun exercise On Tue, Aug 22, 2017 at 2:55 PM, Chris Riley wrote: > The initial message I replied to stated in part, "Okay so I quite like > this idea. If we start removing at height 630000 or 840000 (gives us 4-8 > years to develop this solution), it stays nice and neat with the halving > interval...." > > That is less than 3 years or less than 7 years away. Much sooner than it > is believed QC or Moore's law could impact bitcoin. Changing bitcoin so as > to require that early coins start getting "scavenged" at that date seems > unneeded and irresponsible. Besides, your ECDSA is only revealed when you > spend the coins which does provide some quantum resistance. Hal was just > an example of people putting their coins away expecting them to be there at > X years in the future, whether it is for himself or for his kids and wife. > > :-) > > > > On Tue, Aug 22, 2017 at 1:33 PM, Matthew Beton > wrote: > >> Very true, if Moore's law is still functional in 200 years, computers >> will be 2^100 times faster (possibly more if quantum computing becomes >> commonplace), and so old wallets may be easily cracked. >> >> We will need a way to force people to use newer, higher security wallets, >> and turning coins to mining rewards is better solution than them just being >> hacked. >> >> On Tue, 22 Aug 2017, 7:24 pm Thomas Guyot-Sionnest >> wrote: >> >>> In any case when Hal Finney do not wake up from his 200years >>> cryo-preservation (because unfortunately for him 200 years earlier they did >>> not know how to preserve a body well enough to resurrect it) he would find >>> that advance in computer technology made it trivial for anyone to steal his >>> coins using the long-obsolete secp256k1 ec curve (which was done long >>> before, as soon as it became profitable to crack down the huge stash of >>> coins stale in the early blocks) >>> >>> I just don't get that argument that you can't be "your own bank". The >>> only requirement coming from this would be to move your coins about once >>> every 10 years or so, which you should be able to do if you have your >>> private keys (you should!). You say it may be something to consider when >>> computer breakthroughs makes old outputs vulnerable, but I say it's not >>> "if" but "when" it happens, and by telling firsthand people that their >>> coins requires moving every once in a long while you ensure they won't do >>> stupid things or come back 50 years from now and complain their addresses >>> have been scavenged. >>> >>> -- >>> Thomas >>> >>> >>> On 22/08/17 10:29 AM, Erik Aronesty via bitcoin-dev wrote: >>> >>> I agree, it is only a good idea in the event of a quantum computing >>> threat to the security of Bitcoin. >>> >>> On Tue, Aug 22, 2017 at 9:45 AM, Chris Riley via bitcoin-dev < >>> bitcoin-dev@lists.linuxfoundation.org> wrote: >>> >>>> This seems to be drifting off into alt-coin discussion. The idea that >>>> we can change the rules and steal coins at a later date because they are >>>> "stale" or someone is "hoarding" is antithetical to one of the points of >>>> bitcoin in that you can no longer control your own money ("be your own >>>> bank") because someone can at a later date take your coins for some reason >>>> that is outside your control and solely based on some rationalization by a >>>> third party. Once the rule is established that there are valid reasons why >>>> someone should not have control of their own bitcoins, what other reasons >>>> will then be determined to be valid? >>>> >>>> I can imagine Hal Finney being revived (he was cryo-preserved at Alcor >>>> if you aren't aware) after 100 or 200 years expecting his coins to be there >>>> only to find out that his coins were deemed "stale" so were "reclaimed" (in >>>> the current doublespeak - e.g. stolen or confiscated). Or perhaps he >>>> locked some for his children and they are found to be "stale" before they >>>> are available. He said in March 2013, "I think they're safe enough" stored >>>> in a paper wallet. Perhaps any remaining coins are no longer "safe enough." >>>> >>>> Again, this seems (a) more about an alt-coin/bitcoin fork or (b) better >>>> in bitcoin-discuss at best vs bitcoin-dev. I've seen it discussed many >>>> times since 2010 and still do not agree with the rational that embracing >>>> allowing someone to steal someone else's coins for any reason is a useful >>>> change to bitcoin. >>>> >>>> >>>> >>>> >>>> On Tue, Aug 22, 2017 at 4:19 AM, Matthew Beton via bitcoin-dev < >>>> bitcoin-dev@lists.linuxfoundation.org> wrote: >>>> >>>>> Okay so I quite like this idea. If we start removing at height 630000 >>>>> or 840000 (gives us 4-8 years to develop this solution), it stays nice and >>>>> neat with the halving interval. We can look at this like so: >>>>> >>>>> B - the current block number >>>>> P - how many blocks behind current the coin burning block is. (630000, >>>>> 840000, or otherwise.) >>>>> >>>>> Every time we mine a new block, we go to block (B-P), and check for >>>>> stale coins. These coins get burnt up and pooled into block B's miner fees. >>>>> This keeps the mining rewards up in the long term, people are less likely >>>>> to stop mining due to too low fees. It also encourages people to keep >>>>> moving their money around the enconomy instead of just hording and leaving >>>>> it. >>>>> >>>> >>> > --001a114434fae8f03605575d1f1c Content-Type: text/html; charset="UTF-8" Content-Transfer-Encoding: quoted-printable
> The initial message I replied to stated:
Yes, 3 years is silly.=C2=A0 But coin expiration and quantum resistanc= e is something I've been thinking about for a while, so I tried to stee= r the conversation away from stealing old money for no reason ;).=C2=A0=C2= =A0 Plus I like the idea of making Bitcoin "2000 year proof".
=
- I cannot imagine either SHA256 or any of our existing wallet formats= =20 surviving 200 years, if we expect both moores law and quantum computing to = be a thing.=C2=A0=C2=A0 I would expect the PoW to be rendered obsolete befo= re the Bitcoin addresses.

=C2=A0- A PoW change using Keccak an= d a flexible number of bits can be designed as a "future hard fork&quo= t;.=C2=A0 That is:=C2=A0 the existing POW can be automatically rendered obs= olete... but only in the event that difficulty rises to the level of obsole= scence.=C2=A0=C2=A0 Then the code for a new algorithm with a flexible numbe= r of bits and a difficulty that can scale for thousands of years can then a= utomatically kick in.

=C2=A0- A new addresses format and signing pro= tocols that use a flexible number of bits can be introduced.=C2=A0=C2=A0 Th= e maximum number of supported bits can be configurable, and trivially chang= ed.=C2=A0=C2=A0 These can be made immediately available but completely opti= onal.

=C2=A0- The POW difficulty can be used to inform the expiratio= n of any addresses that can be compromised within 5 years assuming this pow= er was somehow used to compromise them.=C2=A0=C2=A0 Some mechanism for tran= slating global hashpower to brute force attack power can be researched, and= consesrvative estimates made.=C2=A0=C2=A0 Right now, it's like "h= eat death of the universe" amount of time to crack with every machine = on the planet.=C2=A0=C2=A0 But hey... things change and 2000 years is a lon= g time.=C2=A0=C2=A0 This information can be used to inform the expiration a= nd reclamation of old, compromised public addresses.

- Plannin= g a hard fork 100 to 1000 years out is a fun exercise


=

On Tue, Aug 22, 2017 at 2:55 PM, Chris Riley <criley@gmail.com>= wrote:
The initi= al message I replied to stated in part, "Okay so I quite like this ide= a. If we start removing at height 630000 or 840000 (gives us 4-8 years to d= evelop this solution), it stays nice and neat with the halving interval....= "

That is less than 3 years or less than 7 years = =C2=A0away. Much sooner than it is believed QC or Moore's law could imp= act bitcoin.=C2=A0 Changing bitcoin so as to require that early coins start= getting "scavenged" at that date seems unneeded and irresponsibl= e.=C2=A0 Besides, your ECDSA is only revealed when you spend the coins whic= h does provide some quantum resistance.=C2=A0 Hal was just an example of pe= ople putting their coins away expecting them to be there at X years in the = future, whether it is for himself or for his kids and wife. =C2=A0

:-)

<= br>

On = Tue, Aug 22, 2017 at 1:33 PM, Matthew Beton <matthew.beton@gmail.c= om> wrote:

V= ery true, if Moore's law is still functional in 200 years, computers wi= ll be 2^100 times faster (possibly more if quantum computing becomes common= place), and so old wallets may be easily cracked.

We will need a way to force people to use newer, higher secu= rity wallets, and turning coins to mining rewards is better solution than t= hem just being hacked.


On Tue, 22 Aug 2017, 7:24 p= m Thomas Guyot-Sionnest <dermoth@aei.ca> wrote:
=20 =20 =20
In any case when Hal Finney do not wake up from his 200years cryo-preservation (because unfortunately for him 200 years earlier they did not know how to preserve a body well enough to resurrect it) he would find that advance in computer technology made it trivial for anyone to steal his coins using the long-obsolete secp256k1 ec curve (which was done long before, as soon as it became profitable to crack down the huge stash of coins stale in the early blocks)

I just don't get that argument that you can't be "your own= bank". The only requirement coming from this would be to move your coins about once every 10 years or so, which you should be able to do if you have your private keys (you should!). You say it may be something to consider when computer breakthroughs makes old outputs vulnerable, but I say it's not "if" but "when" = it happens, and by telling firsthand people that their coins requires moving every once in a long while you ensure they won't do stupid things or come back 50 years from now and complain their addresses have been scavenged.

--
Thomas


On 22/08/17 10:29 AM, Erik Aronesty via bitcoin-dev wrote:
I agree, it is only a good idea in the event of a quantum computing threat to the security of Bitcoin.=C2=A0=C2=A0

On Tue, Aug 22, 2017 at 9:45 AM, Chris Riley via bitcoin-dev <bitcoin-dev@lists.lin= uxfoundation.org> wrote:
This seems to be drifting off into alt-coin discussion.=C2=A0 The idea that we can change the rules and steal coins at a later date because they are "stale"= ; or someone is "hoarding" is=C2=A0antithetical to one o= f the points of bitcoin in that you can no longer control your own money ("be your own bank") because someone can at a= later date take your coins for some reason that is outside your control and solely based on some rationalization by a third party.=C2=A0 Once the rule is established that there ar= e valid reasons why someone should not have control of their own bitcoins, what other reasons will then be determined to be valid?

I can imagine Hal Finney being revived (he was cryo-preserved at Alcor if you aren't aware) after 100 or 200 years expecting his coins to be there only to find out that his coins were deemed "stale" so we= re "reclaimed" (in the current doublespeak - e.g. st= olen or confiscated).=C2=A0 Or perhaps he locked some for his children and they are found to be "stale" before = they are available.=C2=A0 He said in March 2013, "I think t= hey're safe enough" stored in a paper wallet.=C2=A0 Perhaps a= ny remaining coins are no longer "safe enough."

Again, this seems (a) more about an alt-coin/bitcoin fork or (b) better in bitcoin-discuss at best vs bitcoin-dev. I've seen it discussed many times since 2010 and still do not agree with the rational that embracing allowing someone to steal someone else's coins for any reason is a useful chang= e to bitcoin.




On Tue, Aug 22, 2017 at 4:19 AM, Matthew Beton via bitcoin-dev <bit= coin-dev@lists.linuxfoundation.org> wrote:
Okay so I quite like this idea. If we start removing at height 630000 or 840000 (gives us 4-8 years to develop this solution), it stays nice and neat with the halving interval. We can look at this like so:

B - the current block number
P - how many blocks behind current the coin burning block is. (630000, 840000, or otherwise.)

Every time we mine a new block, we go to block (B-P), and check for stale coins. These coins get burnt up and pooled into block B's miner fees. This keeps the mining rewards up in the long term, people are less likely to stop mining due to too low fees. It also encourages people to keep moving their money around the enconomy instead of just hording and leaving it.



--001a114434fae8f03605575d1f1c--