Return-Path: Received: from smtp1.linuxfoundation.org (smtp1.linux-foundation.org [172.17.192.35]) by mail.linuxfoundation.org (Postfix) with ESMTPS id 2979EAE7 for ; Sun, 28 Jun 2015 20:16:22 +0000 (UTC) X-Greylist: whitelisted by SQLgrey-1.7.6 Received: from mail-ig0-f171.google.com (mail-ig0-f171.google.com [209.85.213.171]) by smtp1.linuxfoundation.org (Postfix) with ESMTPS id 5DA40A7 for ; Sun, 28 Jun 2015 20:16:19 +0000 (UTC) Received: by igcsj18 with SMTP id sj18so65730261igc.1 for ; Sun, 28 Jun 2015 13:16:18 -0700 (PDT) X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:in-reply-to:references:date :message-id:subject:from:to:cc:content-type; bh=aI5pcA657B+OOMcCC44dIU8Z0RvpZ5+Ap7wQYiuTuZ0=; b=hcjnxz+2Kg3UEvMua4jbH8x2SFd3CJtPIfVxdqCdwQwwlHBqru+npVih6XngqdQO9E S6SAhOpZhzBnnF24bFjZUEYc5r71IOhVoaPN+QN6Of5VOYMP4g2O1N46f9ebcnEGOSny gu/IeH3Pfrw22X0H2k799cu+iz55tHjy9bZz6MuOY3Ve0v2/PyipU2tsd89/UXnUWak6 pu1zH3COq4O2TAu7giiDyQ0mP/T1r2mbJsurE1ik0q+jy9e80ItTO4ceSUrGvUwOtCo7 KdMOCdb50/U/yzV9YRY0iDgpapojfKO48UhwQUZgoVmEaVCvyJ1wstXblsdv/9/ePSJF s+Vw== X-Gm-Message-State: ALoCoQkuiO370/6F/MRtwO4e/Hm9toAPjZTuGKJRAoZtnfiW5NvrHWknwWO4mCoOVT+XHMVPTFjj MIME-Version: 1.0 X-Received: by 10.107.3.227 with SMTP id e96mr15807776ioi.50.1435522578754; Sun, 28 Jun 2015 13:16:18 -0700 (PDT) Received: by 10.107.18.97 with HTTP; Sun, 28 Jun 2015 13:16:18 -0700 (PDT) X-Originating-IP: [208.54.4.191] Received: by 10.107.18.97 with HTTP; Sun, 28 Jun 2015 13:16:18 -0700 (PDT) In-Reply-To: <559054D2.3050009@bitcoins.info> References: <558B7352.90708@bitcoins.info> <558D46EC.6050300@bitcoins.info> <558E9C06.9080901@bitcoins.info> <558FF307.9010606@bitcoins.info> <55901F7D.4000001@bitcoins.info> <559054D2.3050009@bitcoins.info> Date: Sun, 28 Jun 2015 13:16:18 -0700 Message-ID: From: Mark Friedenbach To: Milly Bitcoin Content-Type: multipart/alternative; boundary=001a113f0c5a5fc86b051999a5e2 X-Spam-Status: No, score=-2.6 required=5.0 tests=BAYES_00,HTML_MESSAGE, RCVD_IN_DNSWL_LOW autolearn=ham version=3.3.1 X-Spam-Checker-Version: SpamAssassin 3.3.1 (2010-03-16) on smtp1.linux-foundation.org Cc: bitcoin-dev@lists.linuxfoundation.org Subject: Re: [bitcoin-dev] BIP Process and Votes X-BeenThere: bitcoin-dev@lists.linuxfoundation.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Bitcoin Development Discussion List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sun, 28 Jun 2015 20:16:22 -0000 --001a113f0c5a5fc86b051999a5e2 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: quoted-printable Milly you are absolutely wrong as has been pointed out by numerous people numerous times. Your idea of how bitcoin development decision making works is demonstrably false. Please stop filling our inboxes with trolling accusations, or else this will have to become a moderated list. And no one wants that. On Jun 28, 2015 1:11 PM, "Milly Bitcoin" wrote: > I really don't know who has power to do what behind the scenes. From > what i understand, if push comes to shove, it is under the ultimate contr= ol > of one person who can revoke commit privileges. Maybe I am wrong about > that but the point is most people don't know for sure. > > You are correct about the people having the choice to download but the > influence of the official release is way beyond the influence of any fork= ed > release. What that means in the real world is an open question and would > be different depending upon the specific circumstances and difficult to > predict. It is significant power to have control over the official relea= se > at the present time. If they did not have significant power people would > not spend significant efforts lobbying them to make changes. Any new > developers hired by companies will do so because they can influence over > the official release since that is the only incentive. > > It seems to me that this block size fork is only the beginning of the > issues that will arise over the coming years. Whatever powers the core > maintainers have it is going to be exploited one way or another as time > goes on. Maybe there are enough feedback mechanisms to protect against > that, I don't really know. > > Russ > > > > > > On 6/28/2015 3:05 PM, Patrick Murck wrote: > > Wladimir has no more or less =E2=80=9Cpower=E2=80=9D to push change to th= e Bitcoin Core > codebase than any other person with commit privileges to the GitHub repo. > If I=E2=80=99m not mistaken there are 7 people with commit privileges and= five of > them are active. If Wladimir committed a change it could be reverted by a= ny > of the others. This is by design and ensures that changes will have reach= ed > some level of technical consensus before they are merged, among other > things. > > Furthermore even assuming the Core Maintainer commits a change to > Bitcoin Core (that isn=E2=80=99t reverted and that gets packaged up into = the next > code release) that still doesn=E2=80=99t push a change to the bitcoin net= work. > There is no auto-update on Bitcoin Core so individuals and companies > running Bitcoin Core software have to choose to upgrade. Further still, > developers that maintain alternative implementations would have to decide > to merge those changes to the codebase they are indepently maintaining (a= nd > their users would need to update, etc.). > > I understand why it might *seem* to be the case that the Core Maintainer > is empowered to make changes to "teh Bitcoin" but the reality is that the > Core Maintainer role is really about cat herding and project management o= f > Bitcoin Core the open-source software project and not the bitcoin network= . > We=E2=80=99re lucky Wladimir has agreed to take on so much of the scut wo= rk to keep > the project moving forward. > > The process might get ugly and inefficient but that=E2=80=99s the cost o= f having > no wizard behind the curtain. > > -pm > > -- > Patrick Murck > > On June 28, 2015 at 9:23:47 AM, Milly Bitcoin (milly@bitcoins.info) wrote= : > > The core maintainer has always been in control of the consensus rules. > Satoshi came up with the rules and put them in there. Since then any > changes to any part of the code go through the core maintainer. It > looks to me as if people are saying it somehow changed along the way > because they don't want to hurt people's feeling, upset up, get them to > quit, etc. Sure there are checks and balances and people don't have to > use the main code base but if they change the consensus rules they are > incompatible. > > The notion that because people can download different rules and run them > is interesting from a theoretical perspective but that is constrained by > the network effect. I can say the US government is not the "decider" of > laws because I can vote them out, recall them, challenge things in > court, or secede and start my own country. You can also say the > judge/jury in a criminal court case is not a "decider" because the > president can always issue a pardon. But those points are generally not > useful in a practical sense. > > The issue about the developers is the tremendous influence they have to > veto any changes. I don't have veto power yet I have more bitcoins than > garzik says he has. The whole Bitcoin software development system is > subject to attack from just a couple of people who have this veto > power. With all the crying and moaning about centralization on this > list I would think that would be a concern. > > Russ > > > > On 6/28/2015 11:35 AM, Jorge Tim=C3=B3n wrote: > > On Sun, Jun 28, 2015 at 3:13 PM, Milly Bitcoin > wrote: > >> I never said something was approved by garzik added something after it > was > >> opposed. What I said was a proposal was made and 4 people commented on > the > >> Github. He then tweeted there was near universal approval before most > >> people even heard about the subject. It was not controversial but i wa= s > >> pointing out the arrogance of some of the developers. He considers the > >> entire universe of Bitcoin stakeholders to be a very small group of > >> insiders, not the entire universe of Bitcoin users. Another thing I ha= ve > >> seen on Github for bitcoin.org is how some the maintainers change the > rules > >> on the fly. Sometimes they say a proposal had no objections so it is > >> approved. Other times they say a proposal has no support so it is > rejected. > > Ok, I misunderstood. > > Well, the fact is that the number of capable reviewers is quite small. > > If more companies hired and trained more developers to become bitcoin > > core developers that situation could change, but that's where we are > > now. > > > >> You are also trying to say that the core developers actually have litt= le > >> influence and are not "deciders" because anyone can fork the code. Tha= t > has > >> already been discussed at length and such an argument is faulty becaus= e > >> there is a constraint that your software is incompatible with everyone > else. > > Only if you change the consensus rules (which are, in fact, a > > relatively small part of the code). > > Mike mantains Bitcoin XT and that's fine, Peter Todd maintains patches > > with the replace by fee policy, libbitcoin also changes many > > non-consensus things, there's code written in other languages... > > There's multiple counter-examples to your claim of that argument being > faulty. > > Seriously, forking the project is just one click. You should try it > > out like at least 9627 other people have done. > > >From there, you can pay your own developers (if you don't know how to > > code yourself) and maybe they're also fine being insulted by you as > > part of the job. > > What you still can't do is unilaterally change the consensus rules of > > a running p2p consensus system, because you cannot force the current > > users to run any software they don't want to run. > > > >> The issue is that there is no way right now to change the consensus > rules > >> except to go through the core maintainer unless you get everybody on t= he > >> network to switch to your fork. People who keep repeating that the > software > >> development is "decentralized because you fork the code" without > explaining > >> the constraints are just cultists. > > Please, stop the cultist crap. Maybe insulting people like that is how > > you got people to call you a troll. > > But, yes, you are right: there's no known mechanism for safely > > deploying controversial changes to the consensus rules > > > >> The discussion has nothing to do with who has the position now and I > never > >> said he has "control over the consensus rules." The maintainer has a > very > >> large influence way beyond anyone else. As for your claim that I want > >> someone hurt because I am explaining the process, that is ridiculous. = If > >> the Core maintainers did not have significant influence to change the > >> consensus rules then everybody would not be spending all this time > lobbying > >> them to have them changed. > > Well, if you don't think he has control over the consensus rules we're > > advancing. > > I think that was implied from some of your previous claims. He is no > > "decider" on consensus changes. > > Insisting on it can indeed get him hurt, so I'm happy that you're > > taking that back (or clarifying that really wasn't your position). > > Influence is very relative and not only core devs have "influence". > > Maybe Andreas Antonopolous has more "influence" than I have because he > > is a more public figure? > > Well, that's fine I think. I don't see the point in discussing who has > > how much influence. > > > >> The outside influences and stake of the developer is a relevant topic. > The > >> same types of things are discussed on this list all the time in the > context > >> of miners, users, merchants, and exchanges. Again, the developers try = to > >> place themselves on some kind of pedestal where they are the protector= s > and > >> pure and everyone else (miners, users, merchants) are abusers, spammer= s, > >> attackers, scammers, cheaters, etc. It is Garzik who voluntarily made = an > >> issue of how many bitcoins he holds and he made that issue in the same > place > >> where he announces many of the technical issues. It is very relevant > that > >> he has a minimal stake in Bitcoin holdings yet he goes around making > major > >> decisions about Bitcoin and trying to dictate who is allowed to > participate > >> in discussions. If a core developer has minimal stake in Bitcoin yet h= as > >> major veto power over code change that is a problem. > > Please, don't generalize. I don't think I put myself in any kind of > pedestal. > > That is insulting to me and many others (you may not even know and > > you're insulting them). > > And I think my Bitcoin holdings are completely irrelevant when judging > > my contributions to the software: either they're good or not, and who > > I am or how many Bitcoins I have at any given time shouldn't matter. > > Again, nobody forces you to use our software, as said there's > > alternatives (including forking the project right now). > > > >> You are correct that you cannot give power to any person over the > Internet > >> which is why some kind of process needs to be developed that does not > >> involve trying to convince one person to make the changes or a system > that > >> depends on unwritten, ever-changing rules maintained by a handful of > people. > > Well, for now the process we have is seeking consensus, and although > > our definition of "uncontroversial" is very vague, I think it is quite > > obvious when a proposed change is not "uncontroversial" (like in the > > block size debate). > > It seems to me that any other "formal process" would imply > > centralization in the decision making of the consensus rules (and from > > there you only have to corrupt that centralized organization to > > destroy Bitcoin). > > > > > _______________________________________________ > bitcoin-dev mailing list > bitcoin-dev@lists.linuxfoundation.org > https://lists.linuxfoundation.org/mailman/listinfo/bitcoin-dev > > > > _______________________________________________ > bitcoin-dev mailing list > bitcoin-dev@lists.linuxfoundation.org > https://lists.linuxfoundation.org/mailman/listinfo/bitcoin-dev > > --001a113f0c5a5fc86b051999a5e2 Content-Type: text/html; charset=UTF-8 Content-Transfer-Encoding: quoted-printable

Milly you are absolutely wrong as has been pointed out by nu= merous people numerous times. Your idea of how bitcoin development decision= making works is demonstrably false. Please stop filling our inboxes with t= rolling accusations, or else this will have to become a moderated list. And= no one wants that.

On Jun 28, 2015 1:11 PM, "Milly Bitcoin&quo= t; <milly@bitcoins.info> w= rote:
=20 =20 =20
I really don't know who has power to do what behind the scenes.=C2=A0 From what i understand, if push comes t= o shove, it is under the ultimate control of one person who can revoke commit privileges.=C2=A0 Maybe I am wrong about that but the point is most people don't know for sure.

You are correct about the people having the choice to download but the influence of the official release is way beyond the influence of any forked release.=C2=A0 What that means in the real world is an open question and would be different depending upon the specific circumstances and difficult to predict.=C2=A0 It is significant power to have control over the official release at the present time.=C2=A0 = If they did not have significant power people would not spend significant efforts lobbying them to make changes.=C2=A0 Any new developers hired by companies will do so because they can influence over the official release since that is the only incentive.

It seems to me that this block size fork is only the beginning of the issues that will arise over the coming years.=C2=A0 Whatever powe= rs the core maintainers have it is going to be exploited one way or another as time goes on.=C2=A0 Maybe there are enough feedback mechanisms to protect against that, I don't really know.=C2=A0=C2= =A0

Russ





On 6/28/2015 3:05 PM, Patrick Murck wrote:
=20
Wladimir has no more or less =E2=80=9Cpower=E2=80=9D to push change to the Bitco= in Core codebase than any other person with commit privileges to the GitHub repo. If I=E2=80=99m not mistaken there are 7 people with co= mmit privileges and five of them are active. If Wladimir committed a change it could be reverted by any of the others. This is by design and ensures that changes will have reached some level of technical consensus before they are merged, among other things.

Furthermore even assuming the Core Maintainer commits a change to Bitcoin Core (that isn=E2=80=99t reverted and that gets packaged up into th= e next code release) that still doesn=E2=80=99t push a change to the bitcoin network. There is no auto-update on Bitcoin Core so individuals and companies running Bitcoin Core software have to choose to upgrade. Further still, developers that maintain alternative implementations would have to decide to merge those changes to the codebase they are indepently maintaining (and their users would need to update, etc.).

I understand why it might *seem* to be the case that the Core Maintainer is empowered to make changes to "teh Bitcoin" but the realit= y is that the Core Maintainer role is really about cat herding and project management of Bitcoin Core the open-source software project and not the bitcoin network. We=E2=80=99re lucky Wladimir h= as agreed to take on so much of the scut work to keep the project moving forward.

The process might get ugly and inefficient but that=E2=80=99s the cost of havin= g no wizard behind the curtain.

-pm

--=C2=A0
Patrick Murck

On June 28, 2015 at 9:23:47 AM, Milly Bitcoin (milly@bitcoins.info) wrote:

The core maintainer has always been in control of the consensus rules.
Satoshi came up with the rules and put them in there. Since then any
changes to any part of the code go through the core maintainer. It
looks to me as if people are saying it somehow changed along the way
because they don't want to hurt people's feeling, ups= et up, get them to
quit, etc. Sure there are checks and balances and people don't have to
use the main code base but if they change the consensus rules they are
incompatible.

The notion that because people can download different rules and run them
is interesting from a theoretical perspective but that is constrained by
the network effect. I can say the US government is not the "decider" of
laws because I can vote them out, recall them, challenge things in
court, or secede and start my own country. You can also say the
judge/jury in a criminal court case is not a "decider&qu= ot; because the
president can always issue a pardon. But those points are generally not
useful in a practical sense.

The issue about the developers is the tremendous influence they have to
veto any changes. I don't have veto power yet I have more bitcoins than
garzik says he has. The whole Bitcoin software development system is
subject to attack from just a couple of people who have this veto
power. With all the crying and moaning about centralization on this
list I would think that would be a concern.

Russ



On 6/28/2015 11:35 AM, Jorge Tim=C3=B3n wrote:
> On Sun, Jun 28, 2015 at 3:13 PM, Milly Bitcoin <= milly@bitcoins.info> wrote:
>> I never said something was approved by garzik added something after it was
>> opposed. What I said was a proposal was made and 4 people commented on the
>> Github. He then tweeted there was near universal approval before most
>> people even heard about the subject. It was not controversial but i was
>> pointing out the arrogance of some of the developers. He considers the
>> entire universe of Bitcoin stakeholders to be a very small group of
>> insiders, not the entire universe of Bitcoin users. Another thing I have
>> seen on Github for bitcoin.org is how some the maintainers change the rules
>> on the fly. Sometimes they say a proposal had no objections so it is
>> approved. Other times they say a proposal has no support so it is rejected.
> Ok, I misunderstood.
> Well, the fact is that the number of capable reviewers is quite small.
> If more companies hired and trained more developers to become bitcoin
> core developers that situation could change, but that's where we are
> now.
>
>> You are also trying to say that the core developers actually have little
>> influence and are not "deciders" because a= nyone can fork the code. That has
>> already been discussed at length and such an argument is faulty because
>> there is a constraint that your software is incompatible with everyone else.
> Only if you change the consensus rules (which are, in fact, a
> relatively small part of the code).
> Mike mantains Bitcoin XT and that's fine, Peter Todd maintains patches
> with the replace by fee policy, libbitcoin also changes many
> non-consensus things, there's code written in other languages...
> There's multiple counter-examples to your claim of that argument being faulty.
> Seriously, forking the project is just one click. You should try it
> out like at least 9627 other people have done.
> >From there, you can pay your own developers (if you don't know how to
> code yourself) and maybe they're also fine being insulted by you as
> part of the job.
> What you still can't do is unilaterally change the consensus rules of
> a running p2p consensus system, because you cannot force the current
> users to run any software they don't want to run. >
>> The issue is that there is no way right now to change the consensus rules
>> except to go through the core maintainer unless you get everybody on the
>> network to switch to your fork. People who keep repeating that the software
>> development is "decentralized because you fork the code" without explaining
>> the constraints are just cultists.
> Please, stop the cultist crap. Maybe insulting people like that is how
> you got people to call you a troll.
> But, yes, you are right: there's no known mechanism for safely
> deploying controversial changes to the consensus rules
>
>> The discussion has nothing to do with who has the position now and I never
>> said he has "control over the consensus rules.&= quot; The maintainer has a very
>> large influence way beyond anyone else. As for your claim that I want
>> someone hurt because I am explaining the process, that is ridiculous. If
>> the Core maintainers did not have significant influence to change the
>> consensus rules then everybody would not be spending all this time lobbying
>> them to have them changed.
> Well, if you don't think he has control over the consensus rules we're
> advancing.
> I think that was implied from some of your previous claims. He is no
> "decider" on consensus changes.
> Insisting on it can indeed get him hurt, so I'm happ= y that you're
> taking that back (or clarifying that really wasn't your position).
> Influence is very relative and not only core devs have "influence".
> Maybe Andreas Antonopolous has more "influence"= ; than I have because he
> is a more public figure?
> Well, that's fine I think. I don't see the point= in discussing who has
> how much influence.
>
>> The outside influences and stake of the developer is a relevant topic. The
>> same types of things are discussed on this list all the time in the context
>> of miners, users, merchants, and exchanges. Again, the developers try to
>> place themselves on some kind of pedestal where they are the protectors and
>> pure and everyone else (miners, users, merchants) are abusers, spammers,
>> attackers, scammers, cheaters, etc. It is Garzik who voluntarily made an
>> issue of how many bitcoins he holds and he made that issue in the same place
>> where he announces many of the technical issues. It is very relevant that
>> he has a minimal stake in Bitcoin holdings yet he goes around making major
>> decisions about Bitcoin and trying to dictate who is allowed to participate
>> in discussions. If a core developer has minimal stake in Bitcoin yet has
>> major veto power over code change that is a problem.
> Please, don't generalize. I don't think I put my= self in any kind of pedestal.
> That is insulting to me and many others (you may not even know and
> you're insulting them).
> And I think my Bitcoin holdings are completely irrelevant when judging
> my contributions to the software: either they're goo= d or not, and who
> I am or how many Bitcoins I have at any given time shouldn't matter.
> Again, nobody forces you to use our software, as said there's
> alternatives (including forking the project right now).
>
>> You are correct that you cannot give power to any person over the Internet
>> which is why some kind of process needs to be developed that does not
>> involve trying to convince one person to make the changes or a system that
>> depends on unwritten, ever-changing rules maintained by a handful of people.
> Well, for now the process we have is seeking consensus, and although
> our definition of "uncontroversial" is very va= gue, I think it is quite
> obvious when a proposed change is not "uncontroversial" (like in the
> block size debate).
> It seems to me that any other "formal process"= would imply
> centralization in the decision making of the consensus rules (and from
> there you only have to corrupt that centralized organization to
> destroy Bitcoin).
>


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


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

--001a113f0c5a5fc86b051999a5e2--