Return-Path: Received: from smtp1.linuxfoundation.org (smtp1.linux-foundation.org [172.17.192.35]) by mail.linuxfoundation.org (Postfix) with ESMTPS id B2C21A82 for ; Thu, 13 Jul 2017 13:17:16 +0000 (UTC) X-Greylist: whitelisted by SQLgrey-1.7.6 Received: from mail-qt0-f173.google.com (mail-qt0-f173.google.com [209.85.216.173]) by smtp1.linuxfoundation.org (Postfix) with ESMTPS id 1F0AD1E8 for ; Thu, 13 Jul 2017 13:17:15 +0000 (UTC) Received: by mail-qt0-f173.google.com with SMTP id r30so37372009qtc.0 for ; Thu, 13 Jul 2017 06:17:15 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025; h=mime-version:in-reply-to:references:from:date:message-id:subject:to; bh=pXGOSwMxstOQ1zjKjyEzLl0i3nrDB01TschYj373f9M=; b=LbgZoMM2Myqygw4053gHgSfgUBRWHZitj4zuIsdASzA5LxSXacHzOR+OYcIB0LWR05 qnD5Fj/3R3G20jj31HDZlPvBY95AbMxrDIZPW1yF92mH9y34nJ2DTEFMzVDigP2o6Z4L /pW1ASAk0JO+AFQjxsixfj/hdZi1rhYlBuao2MfBCiaWVrrN1x4/3MP6Ntf+ifD2SL2y 5tdpCgi3Dbso8tsZpunkn0FScVJvwbVZ3ELvAv6LnBsFQxEr0BG/5IZFRQ4lOLrMSM3q 5gD0FyUHPty2Z2GJDMOo20MdYdBTXHmXFsBjuvpoaRO9a5ucSPxMg3wfckoZdc+C3k4n rRbw== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:mime-version:in-reply-to:references:from:date :message-id:subject:to; bh=pXGOSwMxstOQ1zjKjyEzLl0i3nrDB01TschYj373f9M=; b=DMdZwqgOWxqS07VMUwWuFJ6O3NCUuJB54ihpetUrsfMX1yh2tOc7TmdgAk68WWIldw KFHCtUrGiUP9UhEyBFbPWEUw9Y7vtNvVy1xFYYmJG2TR7eWNm2MLnaJDzzM1Ya0m2gqu qGb6REQSxs+aY8NpDpOtuEpsK1tarNP+0Ekcaw3r52tpP1Ap1unpdU67JuDnApFJbzd5 OXRyUg2CWuQVfh/Mrgjr+/2IAbKRQnNQPKW+6XDHhOYkvcKRGhmHR41fzIcSlajAZUvQ 97lPR9QRvN/QHmSXSb0mN5c/OlQKTcpDrsrjbi8MgId2ZWByRstigSV+95crM/RnaSQI blZw== X-Gm-Message-State: AIVw111TaLt4WJFWyvVu7q+yhZo2gEaK4miNbmr+48s5bT53cAwSWPoA EBjnfh7BkcPbXXJ2+R53L5owG8Zmfw== X-Received: by 10.200.45.241 with SMTP id q46mr5155534qta.177.1499951834166; Thu, 13 Jul 2017 06:17:14 -0700 (PDT) MIME-Version: 1.0 Received: by 10.12.186.175 with HTTP; Thu, 13 Jul 2017 06:17:13 -0700 (PDT) In-Reply-To: <3277f4ef-a250-d383-8b00-cb912eb19f8b@gmail.com> References: <24f2b447-a237-45eb-ef9f-1a62533fad5c@gmail.com> <83671224-f6ff-16a9-81c0-20ab578aec9d@gmail.com> <6764b8af-bb4c-615d-5af5-462127bbbe36@gmail.com> <117f6a96-6d90-778a-d87a-be72592e31c5@gmail.com> <42C03DFC-C358-4F8C-A088-735910CCC60E@taoeffect.com> <3277f4ef-a250-d383-8b00-cb912eb19f8b@gmail.com> From: =?UTF-8?Q?Hampus_Sj=C3=B6berg?= Date: Thu, 13 Jul 2017 15:17:13 +0200 Message-ID: To: Paul Sztorc , Bitcoin Protocol Discussion Content-Type: multipart/alternative; boundary="001a1137558c4186b3055432c00e" X-Spam-Status: No, score=-1.5 required=5.0 tests=BAYES_00,DKIM_SIGNED, DKIM_VALID, DKIM_VALID_AU, FREEMAIL_FROM, HTML_MESSAGE, RCVD_IN_DNSWL_NONE, RCVD_IN_SORBS_SPAM autolearn=no version=3.3.1 X-Spam-Checker-Version: SpamAssassin 3.3.1 (2010-03-16) on smtp1.linux-foundation.org X-Mailman-Approved-At: Thu, 13 Jul 2017 13:22:39 +0000 Subject: Re: [bitcoin-dev] Drivechain RfD -- Follow Up 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: Thu, 13 Jul 2017 13:17:16 -0000 --001a1137558c4186b3055432c00e Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable In softforks, I would argue that 100% of all nodes and miners need to upgrade to the new rules. This makes sure that trying to incorrectly spend an "AnyOneCanSpend" will result in a hardfork, instead of a temporary (or permanent) chainsplit. With drivechains, it seems like the current plan is to only let the nodes that are interested in the drivechain validate the other chain, and not necessarily 100% of the network. I guess this could be any percentage of the network, which could lead to a temporary/permanent chainsplit depending on how many percentage of the miners are also validating the other chain (am I missing something here?). I have no way to evaluate if this is an okay trade-off. It seems like major disruption could very likely happen if say only 5% of all fullnodes validate the drivechain. To be fully secure, it seems like 100% of all nodes should also have a fullnode for the drivechain as well... This is one of the reasons I don't advocate sidechains/drivechains as a scaling solution, it looks like it would have to the same outcome as a blocksize increase on the mainchain, but with more complexity. I think sidechains/drivechains could be useful for other things though. Thanks for all your work so far Paul. Hampus 2017-07-13 4:58 GMT+02:00 Paul Sztorc via bitcoin-dev < bitcoin-dev@lists.linuxfoundation.org>: > I will repeat that Drivechain can sometimes be confusing because it is > different things to different people. > > Here is my attempt to break it down into different modes: > > [DC#0] -- Someone who does not upgrade their Bitcoin software (and is > running, say, 0.13). However, they experience the effects of the new rule= s > which miners add (as per the soft fork[s] to add drivechain functionality > and individual drivechains). > [DC#1] -- Someone who always upgrades to the latest version of the Bitcoi= n > software, but otherwise has no interest in running/using sidechains. > [DC#2] -- Someone who upgrades to the latest Bitcoin version, and decides > to also become a full node of one or more sidechains, but who ever actual= ly > uses the sidechains. > [DC#3] -- Someone who upgrades their software, runs sidechain full nodes, > and actively moves money to and from these. > > Greg is still conflating modes [DC#1] and [DC#3]. Specifically, he > equivocates on the team "steal", using it to mean two different things: [= a] > spending an invalid transaction, and [b] a withdrawal that would not matc= h > the report given by a sidechain node. > > The two are quite different, see my comments below: > > > On 7/12/2017 9:15 PM, Tao Effect wrote: > > FYI that document is nearly two years old, and although it is still > overwhelmingly accurate, new optimizations allow us (I think) to push the > waiting period to several weeks and the total ACK counting period up to > several months. > > What does that have to do with my question? The counting period, if I > understood correctly, is something miners do, not full nodes. > > > Greg quoted a passage that contained an older parameter estimate of "a fe= w > days", and I thought it would be helpful and informative if I clarified > that the parameter estimate had been updated to a new (more secure) value= . > > In point of fact, he is wrong, because nodes do the counting. When miners > find a block, they can choose to move the counter up, down, or not at all= . > But nodes do the counting. > > > Also, if the document is old and/or outdated, do you happen to have a lin= k > to a more update-to-date version of the spec? It would be helpful for > review. > > > As I stated above, the document is mostly accurate. There is no other mor= e > up to date version. > > > Because if a node doesn't have the sidechain's information, it will just > assume every withdrawal is valid. This is comparable to someone who still > hasn't upgraded to support P2SH, in cases [DC#0] and [#1]. > > > Right, for [DC#0] and [DC#1], but for [DC#2] an [DC#3] you marked it as > "Yes" without substantiating why you did so. > > > > Above, Greg omitted his original question. For reference, it was: > > The Drivechain spec seems to claim that its use of anyone-can-pay is the= same as P2SH > > > The answer is that both DC and P2SH use transactions which are 'always > valid' to some group of people (un-upgraded users) but which are sometime= s > invalid to new users. So the only difference would be for [DC#0] vs other > versions, but this difference is trivial as miners will censor invalid tx= ns. > > It is your classic soft fork situation. > > > Again, from the perspective of a mainchain user, every withdrawal is vali= d. > > And that is not how P2SH works. > > > Again, keep in mind that Greg continually conflates two different things: > 1. Whether the soft fork rules have been followed, and > 2. Whether the WT^ submitted by a majority hashrate matches the one > calculated by sidechain nodes. > > The first case is exactly equal to P2SH. Just as old nodes accept every > P2SH transaction, so too will [DC#0] users accept every WT^ transaction. > > In the second case, it so happens that [DC#1], [DC#2], and [DC#3] would > also accept any WT^ *that followed the Drivechain rules*, even if they di= d > not like the outcome (because the outcome in question was arbitrarily > designated as a "theft" of funds -- again, see the second case in the lis= t > above). In this way, it is exactly similar to P2SH because nodes will > accept *any* p2sh txn **that follows the p2sh rules**, even if they don't > "like" the specific script contained within (for example, because it is a > theft of "their" BitFinex funds, or a donation to a political candidate > they dislike, etc). > > > [DC#2] and [DC#3] would certainly have an interest in understanding what > is going on, but that has absolutely nothing whatsoever to do with Bitcoi= n > Core and so is off-topic for this mailing list. > > How is that an answer to my question? > > > Greg is, of course, not entitled to an answer to irrelevant questions -- > just as he would not be entitled to an answer if he asked for my favorite > color or my home address. And such answers would needlessly consume the > mailing list's scarce time. > > > What does "[DC#2] and [DC#3] would certainly have an interest in > understanding what is going on" mean? > > > It is clear to me that, if we are not clear on the basics first, we canno= t > hope to tackle anything intermediate. We will come back to this after > clearing up soft fork part. > > > In P2SH, all upgraded nodes will reject invalid P2SH transactions, period= . > > > In DC, all upgraded nodes will reject invalid DC transactions, period. > > > What exactly would [DC#2] and [DC#3] nodes do when faced with an invalid > WT^ transaction =E2=80=94 invalid in the sense that it contains funds whi= ch miners > are stealing? > > > The [DC#2] and [DC#3] nodes would do exactly what the [DC#0] and [DC#1] > nodes do. This is what I mean by "every withdrawal is valid". > > > Again, in P2SH miners cannot steal funds, because all full nodes have a > fully automatic enforcement policy. > > > In DC, miners cannot steal funds, because all full nodes have a fully > automatic enforcement policy. > > However, DC *allows* users to choose to place some of their BTC at the > relative mercy of the miners in creative ways, if they wish (as does P2SH > -- someone could write a script which donates funds to miners, and then > fund it... "paying" to that script). This is another example of conflatin= g > [DC#1] and [DC#3]. > > Paul > > > > > On Jul 12, 2017, at 5:26 PM, Paul Sztorc wrote: > > The confusion below stems from his conflation of several different ideas. > > I will try to explicitly clarify a distinction between several types of > user (or, "modes" of use if you prefer): > > [DC#0] -- Someone who does not upgrade their Bitcoin software (and is > running, say, 0.13). However, they experience the effects of the new rule= s > which miners add (as per the soft fork[s] to add drivechain functionality > and individual drivechains). > [DC#1] -- Someone who always upgrades to the latest version of the Bitcoi= n > software, but otherwise has no interest in running/using sidechains. > [DC#2] -- Someone who upgrades to the latest Bitcoin version, and decides > to also become a full node of one or more sidechains, but who ever actual= ly > uses the sidechains. > [DC#3] -- Someone who upgrades their software, runs sidechain full nodes, > and actively moves money to and from these. > > > On 7/12/2017 6:43 PM, Tao Effect wrote: > > > I am now looking closer again at step number 4 in the Drivechain > specification [2]: > > 4. Everyone waits for a period of, say, 3 days. This gives everyone an > opportunity to make sure the same WT^ is in both the Bitcoin coinbase and > the Sidechain header. If they=E2=80=99re different, everyone has plenty o= f time to > contact each other, figure out what is going on, and restart the process > until its right. > > It seems to me that where our disagreement lies is in this point. > The Drivechain spec seems to claim that its use of anyone-can-pay is the > same as P2SH (and in later emails you reference SegWit as well). Is this > really true? > > FYI that document is nearly two years old, and although it is still > overwhelmingly accurate, new optimizations allow us (I think) to push the > waiting period to several weeks and the total ACK counting period up to > several months. > > [DC#0] Yes > [DC#1] Yes > [DC#2] Yes > [DC#3] Yes > > Because if a node doesn't have the sidechain's information, it will just > assume every withdrawal is valid. This is comparable to someone who still > hasn't upgraded to support P2SH, in cases [DC#0] and [#1]. > > (And this is the main advantage of DC over extension blocks). > > > 2. Per the question in [1], it's my understanding that P2SH transactions > contain all of the information within themselves for full nodes to act as= a > check on miners mishandling the anyone-can-spend nature of P2SH > transactions. However, that does not seem to be the case with WT^ > transactions. > > [DC#0] They do. > [DC#1] They do. > [DC#2] They do. > [DC#3] They do. > > Again, from the perspective of a mainchain user, every withdrawal is vali= d. > > > In P2SH txns, there is no need for anyone to, as the Drivechain spec says= , > "to contact each other, figure out what is going on". Everything just > automatically works. > > There is no *need* to this in Drivechain, either, for [DC#0] or [DC#1]. > > [DC#2] and [DC#3] would certainly have an interest in understanding what > is going on, but that has absolutely nothing whatsoever to do with Bitcoi= n > Core and so is off-topic for this mailing list. > > > If the security of WT^ transactions could be brought up to actually be in > line with the security of P2SH and SegWit transactions, then I would have > far less to object to. > > Somehow I doubt it. > > > Paul > > > > > _______________________________________________ > bitcoin-dev mailing list > bitcoin-dev@lists.linuxfoundation.org > https://lists.linuxfoundation.org/mailman/listinfo/bitcoin-dev > > --001a1137558c4186b3055432c00e Content-Type: text/html; charset="UTF-8" Content-Transfer-Encoding: quoted-printable
In softforks, I would argue that 100% of all no= des and miners need to upgrade to the new rules.
This makes sure that tr= ying to incorrectly spend an "AnyOneCanSpend" will result in a ha= rdfork, instead of a temporary (or permanent) chainsplit.

With= drivechains, it seems like the current plan is to only let the nodes that = are interested in the drivechain validate the other chain, and not necessar= ily 100% of the network.
I guess this could be any percentage= of the network, which could lead to a temporary/permanent chainsplit depen= ding on how many percentage of the miners are also validating the other cha= in (am I missing something here?).

I have no way to evaluate if this= is an okay trade-off.
It seems like major disruption could very likely = happen if say only 5% of all fullnodes validate the drivechain.

To be fully secure, it seems like 100% of all nodes should also have= a fullnode for the drivechain as well...
This is one of the = reasons I don't advocate sidechains/drivechains as a scaling solution, = it looks like it would have to the same outcome as a blocksize increase on = the mainchain, but with more complexity.
I think sidechains/d= rivechains could be useful for other things though.


T= hanks for all your work so far Paul.
Hampus

2017-07-13 4:58 GMT= +02:00 Paul Sztorc via bitcoin-dev <bitcoin-dev@lists.= linuxfoundation.org>:
=20 =20 =20
I will repeat that= Drivechain can sometimes be confusing because it is different things to different people.

Here is my attempt to break it down into different modes:

[DC#0] -- Someone who does not upgrade their Bitcoin software (and is running, say, 0.13). However, they experience the effects of the new rules which miners add (as per the soft fork[s] to add drivechain functionality and individual drivechains).
[DC#1] -- Someone who always upgrades to the latest version of the Bitcoin software, but otherwise has no interest in running/using sidechains.
[DC#2] -- Someone who upgrades to the latest Bitcoin version, and decides to also become a full node of one or more sidechains, but who ever actually uses the sidechains.
[DC#3] -- Someone who upgrades their software, runs sidechain full nodes, and actively moves money to and from these.

Greg is still conflating modes [DC#1] and [DC#3]. Specifically, he equivocates on the team "steal", using it to mean two diffe= rent things: [a] spending an invalid transaction, and [b] a withdrawal that would not match the report given by a sidechain node.

The two are quite different, see my comments below:<= br>

On 7/12/2017 9:15 PM, Tao Effect wrote:
FYI that document is nearly two years old, and although it is still overwhelmingly accurate, new optimizations allow us (I think) to push the waiting period to several weeks and the total ACK counting period up to several months.
What does that have to do with my question? The counting period, if I understood correctly, is something miners do, not full nodes.

Greg quoted a passage that contained an older parameter estimate of "a few days", and I thought it would be helpful and informati= ve if I clarified that the parameter estimate had been updated to a new (more secure) value.

In point of fact, he is wrong, because nodes do the counting. When miners find a block, they can choose to move the counter up, down, or not at all. But nodes do the counting.


Also, if the document is old and/or outdated, do you happen to have a link to a more update-to-date version of the spec? It would be helpful for review.

As I stated above, the document is mostly accurate. There is no other more up to date version.


Because if a node doesn't have the sidechain's information, it will = just assume every withdrawal is valid. This is comparable to someone who still hasn't upgraded to support P2SH, in cases [DC#0] and [#1].

Right, for [DC#0] and [DC#1], but for [DC#2] an [DC#3] you marked it as "Yes" without substantiating why = you did so.


Above, Greg omitted his original question. For reference, it was:

 The Drivechain spec seems to claim that its use of anyone-can-pay is th=
e same as P2SH

The answer is that both DC and P2SH use transactions which are 'always valid' to some group of people (un-upgraded users) but = which are sometimes invalid to new users. So the only difference would be for [DC#0] vs other versions, but this difference is trivial as miners will censor invalid txns.

It is your classic soft fork situation.


Again, from the perspective of a mainchain user, every withdrawal is valid.
And that is not how P2SH works.

Again, keep in mind that Greg continually conflates two different things:
1. Whether the soft fork rules have been followed, and
2. Whether the WT^ submitted by a majority hashrate matches the one calculated by sidechain nodes.

The first case is exactly equal to P2SH. Just as old nodes accept every P2SH transaction, so too will [DC#0] users accept every WT^ transaction.

In the second case, it so happens that [DC#1], [DC#2], and [DC#3] would also accept any WT^ *that followed the Drivechain rules*, even if they did not like the outcome (because the outcome in question was arbitrarily designated as a "theft" of funds -- again, se= e the second case in the list above). In this way, it is exactly similar to P2SH because nodes will accept *any* p2sh txn **that follows the p2sh rules**, even if they don't "like" the specific scri= pt contained within (for example, because it is a theft of "their&quo= t; BitFinex funds, or a donation to a political candidate they dislike, etc).


[DC#2] and [DC#3] would certainly have an interest in understanding what is going on, but that has absolutely nothing whatsoever to do with Bitcoin Core and so is off-topic for this mailing list.
How is that an answer to my question?

Greg is, of course, not entitled to an answer to irrelevant questions -- just as he would not be entitled to an answer if he asked for my favorite color or my home address. And such answers would needlessly consume the mailing list's scarce time.


What does "[DC#2] and [DC#3] would certainly have an interest in understanding what is going on" mean?

It is clear to me that, if we are not clear on the basics first, we cannot hope to tackle anything intermediate. We will come back to this after clearing up soft fork part.


In P2SH, all upgraded nodes will reject invalid P2SH transactions, period.

In DC, all upgraded nodes will reject invalid DC transactions, period.


What exactly would [DC#2] and [DC#3] nodes do when faced with an invalid WT^ transaction =E2=80=94 invalid in the sens= e that it contains funds which miners are stealing?

The [DC#2] and [DC#3] nodes would do exactly what the [DC#0] and [DC#1] nodes do. This is what I mean by "every withdrawal is valid= ".


Again, in P2SH miners cannot steal funds, because all full nodes have a fully automatic enforcement policy.

In DC, miners cannot steal funds, because all full nodes have a fully automatic enforcement policy.

However, DC *allows* users to choose to place some of their BTC at the relative mercy of the miners in creative ways, if they wish (as does P2SH -- someone could write a script which donates funds to miners, and then fund it... "paying" to that script). This is another example of conflating [DC#1] and [DC#3].=

Paul




On Jul 12, 2017, at 5:26 PM, Paul Sztorc <truthcoin@gmail.com> wrote:

=20
The co= nfusion below stems from his conflation of several different ideas.

I will try to explicitly clarify a distinction between several types of user (or, "modes" of use if yo= u prefer):

[DC#0] -- Someone who does not upgrade their Bitcoin software (and is running, say, 0.13). However, they experience the effects of the new rules which miners add (as per the soft fork[s] to add drivechain functionality and individual drivechains).
[DC#1] -- Someone who always upgrades to the latest version of the Bitcoin software, but otherwise has no interest in running/using sidechains.
[DC#2] -- Someone who upgrades to the latest Bitcoin version, and decides to also become a full node of one or more sidechains, but who ever actually uses the sidechains.
[DC#3] -- Someone who upgrades their software, runs sidechain full nodes, and actively moves money to and from these.


On 7/12/2017 6:43 PM, Tao Effect wrote:
=20 =20

I am now looking closer again at step number 4 in the Drivechain specification [2]:

4. Everyone waits for a period of, say, 3 days. This gives everyone an opportunity to make sure the same WT^ is in both the Bitcoin coinbase and the Sidechain header. If they=E2=80=99= re different, everyone has plenty of time to contact each other, figure out what is going on, and restart the process until its right.
It seems to me that where our disagreement lies is in this point.
The Drivechain spec seems to claim that its use of anyone-can-pay is the same as P2SH (and in later emails you reference SegWit as well). Is this really true?
FYI that document is nearly two years old, and although it is still overwhelmingly accurate, new optimizations allow us (I think) to push the waiting period to several weeks and the total ACK counting period up to several months.

[DC#0] Yes
[DC#1] Yes
[DC#2] Yes
[DC#3] Yes

Because if a node doesn't have the sidechain's information, it will just assume every withdrawal is valid. This is comparable to someone who still hasn't upgraded to support P2SH, in cases [DC#0] and [#1].

(And this is the main advantage of DC over extension blocks).


2. Per the question in [1], it's my understanding that P2SH transactions contain all of the information within themselves for full nodes to act as a check on miners mishandling the anyone-can-spend nature of P2SH transactions. However, that does not seem to be the case with WT^ transactions.
[DC#0] They do.
[DC#1] They do.
[DC#2] They do.
[DC#3] They do.

Again, from the perspective of a mainchain user, every withdrawal is valid.


In P2SH txns, there is no need for anyone to, as the Drivechain spec says, "to contac= t each other, figure out what is going on". Everythi= ng just automatically works.
There is no *need* to this in Drivechain, either, for [DC#0] or [DC#1].

[DC#2] and [DC#3] would certainly have an interest in understanding what is going on, but that has absolutely nothing whatsoever to do with Bitcoin Core and so is off-topic for this mailing list.


If the security of WT^ transactions could be brought up to actually be in line with the security of P2SH and SegWit transactions, then I would have far less to object to.
Somehow I doubt it.


Paul



_______________________________________________
bitcoin-dev mailing list
bitcoin-dev@lists.= linuxfoundation.org
https://lists.linuxfoundation.org= /mailman/listinfo/bitcoin-dev


--001a1137558c4186b3055432c00e--