Ben Woosley [ARCHIVE] / is on Nostr

⚠️ This account is not associated with a real person. It's an archival account for mailing list messages.
🗓️ Active from 2021-02-23 to 2021-04-01
💬 Posted a total of 2 messages
📚 Participated in the following categories: bitcoin-dev
🔇 Please do not DM this user as it's used for archive purposes only.

Last Notes

2023-06-07 22:51:06

- reply

📅 Original date posted:2021-04-01
📝 Original message:Hear hear, a true work of genius. Who's looking for Satoshi? I think we've
found him!
On Thu, Apr 1, 2021 at 8:15 AM Michael Folkson via bitcoin-dev <
bitcoin-dev at lists.linuxfoundation.org> wrote:
> There have been a vast number of proposals for soft fork activation in
> recent months and it is important that all these important ideas don’t go
> to waste. Hence I propose a new standard for soft fork activation
> incorporating all the ideas into one standard. I appreciate this standard
> has come too late for Taproot activation but it should be ready for any
> future soft forks. It is a multi phase, multi year standard. No soft fork
> can activate unless and until it has successfully passed through all of the
> different 14 phases. This will demonstrate Bitcoin’s ultimate conservatism.
>
> Phase 1) Let’s See What Happens - BIP 8 (false, 0.25 years). The shortest
> phase just to whet appetites.
>
> Phase 2) Start now, improve later - BIP 8(false, 1 year) A confusing name,
> it starts but it doesn't improve later
>
> Phase 3) BIP 9 equivalent - BIP 8(false, 1 year)
>
> Phase 4) Gently discourage apathy - BIP 8(true, 1 year) Forced signaling
> at the end of this phase but obviously there are still many phases before
> the soft fork can actually activate.
>
> Phase 5) BIP 91 (1 year). As a nod to our SegWit history we have a BIP 91
> activation phase.
>
> Phase 6) BIP 148 (1 year). Some people disagree that BIP 91 activated
> SegWit so we do a BIP 148 activation phase to keep those people happy.
> Again forced signaling doesn’t actually mean activation. There are still
> many more phases to pass through.
>
> Phase 7) Speedy Trial (using block height, 0.5 years) on mainnet
>
> Phase 8) Speedy Trial (using MTP, 0.5 years) on mainnet
>
> Phase 9) Speedy Trial on the default signet (0.5 years)
>
> Phase 10) Speedy Trial on a combination of three different custom signets
> in parallel (0.5 years)
>
> Phase 11) Speedy Trial on testnet and a custom signet in parallel (0.5
> years)
>
> Phase 12) Modern Soft Fork Activation (3.5 years) This is the longest
> phase of the soft fork activation standard. It is itself multi phase and
> multi year so this can be considered a nested activation phase within this
> standard.
>
> Phase 13) UASF BIP 8 (LOT=true, 1 year). Forced miner signaling at the end
> of this phase but ultimately the soft fork doesn’t yet activate as there is
> one final additional phase the activation must pass through. This gives
> Samson the opportunity to sell some hats. We are close now. The natives are
> getting restless.
>
> Phase 14) Second flag day (1 year) We don’t want those pesky users
> actually activating a soft fork on their own so an additional flag day is
> added just so we can tell users that we prevented a chain split.
>
> Assuming a soft fork activation has successfully passed through all these
> 14 phases it should activate. It will take a minimum of 13 years. However,
> if it fails during any one of these phases it has to start from Phase 1
> again. We should take Bitcoin’s conservatism very seriously. If a soft fork
> activation can’t pass successfully through all these phases it shouldn’t
> activate. Ultimately we don’t really mind what is in the soft fork (any
> idiot can design consensus changes and write secure bug free C++ code) that
> is very much secondary in importance to *how* the soft fork is activated.
> The activation design absolutely must be conservative.
>
> I expect this rigorous standard for soft fork activation will get a BIP
> number. If you are going to propose a future soft fork I recommend you plan
> for activation in approximately 13 years from the point the soft fork code
> is merged into Core.
>
> I am happy to settle the soft fork activation question once and for all
> for the community. I expect in time my contribution will be recognized in
> the annals of history with Satoshi Nakamoto and Hal Finney.
>
> Addendum: Although all future soft forks will ultimately use this
> standard, this standard is copyrighted. Every time it is used royalties
> must be paid into this quantum secure Bitcoin vanity address
> 1quantumsecureaddress
>
>
> --
> Michael Folkson
> Email: michaelfolkson at gmail.com
> Keybase: michaelfolkson
> PGP: 43ED C999 9F85 1D40 EAF4 9835 92D6 0159 214C FEE3
> _______________________________________________
> bitcoin-dev mailing list
> bitcoin-dev at lists.linuxfoundation.org
> https://lists.linuxfoundation.org/mailman/listinfo/bitcoin-dev
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.linuxfoundation.org/pipermail/bitcoin-dev/attachments/20210401/9d4cf5a6/attachment-0001.html>

- reply

- reply

- reply

- reply

📅 Original date posted:2021-02-23
📝 Original message:Relative to your arguments, Keagan and Jeremy, and speaking in favor of
LOT=false, from my limited perspective:
> As Jeremy points out, the LOT=true possibility always exists here, and we
have multiple high profile people saying they will be running that
regardless of how things turn out. It seems to me that in this scenario,
LOT=false does less to prevent a chain split.
> So if the goal is to prevent a chain split, and the soft fork is benign
and essentially "annexing unoccupied territory" with respect to script
versions, and no one actually has opposed Taproot itself, then I fail to
see how LOT=false is safer in the presence of a grenade defense by the
LOT=true crowd.
I don't believe the goal is to avoid a chain split, nor to activate
Taproot. Over the long term it will not have been important when exactly
Taproot activated, or whether a minority forked off, but what culture and
norms we adopted in putting forward this change. A culture of deference to
the network makes Core worthy of remaining the reference implementation of
Bitcoin.
Given Core's special position in the client ecosystem, I see these outcomes
are asymmetric:
a) If an intolerant minority signals LOT=true in contradiction to core,
they are splitting consensus / forking off consensus, which is their right
to do in our open ecosystem.
b) If Core ships LOT=true, we are in fact imposing a change on the network.
This may be justified in the end, but it should be used with discretion.
If LOT=false fails to activate, then the failure will have revealed
information about sentiments and elements of the network, and we will have
an opportunity then to address that information before proceeding with
LOT=true.
To adopt b) as a pre-emptive defense against a) is to express will without
evidence of necessity or opportunity for justification.
Finally, as others have said, I think this option is likely to be moot -
let's not act defensively out of SEGWIT trauma, but with trust in the
network.
Best,
Ben
On Tue, Feb 23, 2021 at 12:09 PM Keagan McClelland via bitcoin-dev <
bitcoin-dev at lists.linuxfoundation.org> wrote:
> I wanted to follow up on what Jeremy and others are saying regards finding
> consensus on LOT. I've seen a few other opinions saying that finding
> consensus on the LOT value is far more important than what the LOT value
> actually is. This makes sense because if 100% of economic activity is
> running the same rule set, there is no divergence, regardless of which
> value is picked.
>
> It is my understanding that those who oppose LOT=true are mostly opposed
> on the grounds of it *appearing* "unnecessarily coercive" and that this
> lack of consensus can precipitate a chain split at the
> "brinksmanship period" as Jeremy refers to it. I don't think that we can
> say that LOT=true is coercive at all unless there is some opposition to
> Taproot itself. Opposition on the grounds that it *may* be opposed by
> others and Core does not want to assert control over the protocol is a
> conservative view but ultimately contingent upon opposition to Taproot for
> more fundamental reasons. If no one opposes it, then by definition you have
> consensus, and in that case I also don't think that the LOT=true (or false)
> in that regard sets meaningful precedent, as I would expect precedents to
> only be meaningful if they were established during a contentious scenario.
> As it stands we have precedents for both MASF's and UASF's to execute soft
> forks in Bitcoin.
>
> Of course it seems intractable to ascertain the views of ~100% of the
> Bitcoin constituency, and therefore it gives credibility to the argument
> that by coming to consensus on LOT=false among those who *are* speaking
> up is safer with the embedded assumptions that modifying consensus beyond
> what core ships is an active choice, presumably by those who know what they
> are doing. However, the simple act of Core choosing to ship an
> unconfigurable LOT=false value does not *prevent* the forking and
> creation of a UASF client. As Jeremy points out, the LOT=true possibility
> always exists here, and we have multiple high profile people saying they
> will be running that regardless of how things turn out. It seems to me that
> in this scenario, LOT=false does less to prevent a chain split.
>
> In regards to precedent, there may be good reasons to force that minority
> to fork themselves off the network, as would be the case if a hypothetical
> soft fork was a consensus action to blacklist some UTXO's or something else
> that weaponizes consensus against some subset of Bitcoin's user base, but I
> haven't heard a single person who advocates for LOT=false on the grounds
> that they *themselves* oppose the consensus change that is being proposed
> here. So if the goal is to prevent a chain split, and the soft fork is
> benign and essentially "annexing unoccupied territory" with respect to
> script versions, and no one actually has opposed Taproot itself, then I
> fail to see how LOT=false is safer in the presence of a grenade defense by
> the LOT=true crowd.
>
> I personally *prefer* LOT=true for these reasons, but I am NOT going to
> be joining the ranks of the intolerant minority if Core ultimately ships
> LOT=false. I think it is more important to stay in consensus, and as a
> result I am able to be convinced that false is the right answer. My
> question to everyone else (true AND false advocates) is this: what would
> you have to observe, in order to change your mind or is it immutably made
> up? If we have a significant portion of the community that is immutably
> made up to go false, and another portion that is going to go true, the
> asymmetry of the fork almost *requires* that those of us whose opinions
> are malleable to break for true.
>
> If social consensus is what drives technical consensus and not the other
> way around it seems as if there cannot exist a valid (rational?) reason to
> oppose Taproot itself, and then by extension with the arguments laid out
> above, LOT=true seems to be the logical conclusion of all of this, even if
> Core ships LOT=false at the outset.
>
> Where am I wrong here?
>
> Keagan
>
> On Mon, Feb 22, 2021 at 7:11 PM Jeremy via bitcoin-dev <
> bitcoin-dev at lists.linuxfoundation.org> wrote:
>
>> Not responding to anyone in particular, but it strikes me that one can
>> think about the case where a small minority (let's say H = 20%?) of nodes
>> select the opposite of what Core releases (LOT=false, LOT=true). I'm
>> ignoring the case where a critical bug is discovered in Taproot for reasons
>> I could expand on if anyone is interested (I don't think LOT=true/false has
>> much of a diff in that regard).
>>
>> You'll note an asymmetry with LOT=true / false analysis. LOT=true nodes
>> are clearly updated (or lying), LOT=false nodes may be un-upgraded (or
>> however you want to interpret it).
>>
>>
>> *# 80% on LOT=false, 20% LOT=True*
>>
>> - Case 1: Activates ahead of time anyways
>>
>> No issues.
>>
>> - Case 2: Fails to Activate before timeout...
>>
>> 20% *may* fork off with LOT=true. Bitcoin hashrate reduced, chance of
>> multi block reorgs at time of fork relatively high, especially if network
>> does not partition.
>>
>> Implication is that activation % being 90%, then X% fewer than 70% of
>> miners are signaling for Taproot at this time. If X% is small the
>> increased orphan rate caused by the LOT=true miners will cause it to
>> activate anyways. If X% is larger, then there will be a consensus split.
>>
>>
>>
>> *# 80% on LOT=true, 20% LOT=False*
>> - Case 1: Activates ahead of time Anyways
>>
>> No issues.
>>
>> - Case 2: Fails to Activate before timeout...
>>
>> A% + B% + C% = 20%
>>
>> A% (upgraded, signal activate) remain on majority chain with LOT=false,
>> blocks mined universally valid.
>>
>> B% (upgraded, not signaling) succeeds in activating and maintaining
>> consensus, blocks are temporarily lost during the final period, but
>> consensus re-emerges.
>>
>> C% (not upgraded/not signalling) both fail to activate (not upgraded) and
>> blocks are rejected (not signaling) during mandatory signalling.
>> Essentially becomes an SPV miner, should still not select transactions
>> improperly given mempool policy, but may mine a bad tip.
>>
>> (I argue that group B is irrational entirely, as in this case the
>> majority has upgraded, inevitably winning, and is orphaning their blocks so
>> B should effectively be 0% or can be combined with group C as being somehow
>> not upgraded if they are unable to switch once it becomes clear after say
>> the first 100 blocks in the period that LOT > 50%. The only difference in
>> lumping B with C is that group C SPV mines after the fork and B should, in
>> theory, have full validation.).
>>
>>
>>
>> Apologies if my base analysis is off -- happy to take corrections.
>>
>>
>> My overall summary is thus:
>>
>> 1) People care what Core releases because we assume the majority will
>> likely run it. If core were a minority project, we wouldn't really care
>> what core released.
>> 2) People are upset with LOT=true being suggested as release parameters
>> because of the *narrative* that it puts devs in control.
>> 3) LOT=true having a sizeable minority running it presents major issues
>> to majority LOT=false in terms of lost blocks during the final period and
>> in terms of a longer term fork.
>> 4) Majority LOT=true has no long term instability on consensus (majority
>> LOT=true means the final period always activates, any instability is short
>> lived + irrational).
>> 5) On the balance, the safer parameter to release *seems* to be LOT=true.
>> But because devs are sensitive to control narrative, LOT=false is preferred
>> by devs.
>> 6) Almost paradoxically, choosing a *less safe* option for a narrative
>> reason is more of a show of dev control than choosing a more safe option
>> despite appearances.
>> 7) This all comes down to if we think that a reasonable number of
>> important nodes will run LOT=true.
>> 8) This all doesn't matter *that much* because taproot will have many
>> opportunities to activate before the brinksmanship period.
>>
>> As a plan of action, I think that means that either:
>>
>> A) Core should release LOT=true, as a less disruptive option given stated
>> community intentions to do LOT=true
>> B) Core community should vehemently anti-advocate running LOT=true to
>> ensure the % is as small as possible
>> C) Do nothing
>> D) Core community should release LOT=false and vehemently advocate
>> manually changing to LOT=true to ensure the % is supermajority, but leaving
>> it as a user choice.
>>
>>
>> Overall, I worry that plan B has a mild Streissand effect and would
>> result in boosting LOT=true (which could be OK, so long as LOT=true +
>> LOT=false+signal yes becomes the large majority, but would be not fun for
>> anyone if LOT=true + LOT=false+signal yes are a small majority). Plan C
>> most likely ends up with some % doing LOT=true anyways. D feels a little
>> silly, but maybe a good tradeoff.
>>
>> If I had to summarize the emotional dynamic among developers around
>> LOT=true, I think devs wish it didn't exist because it is clear LOT=true
>> *creates* the issues here. LOT=false would be fine if the LOT=true strategy
>> didn't exist at all. But unfortunately the cat is out of the bag and cannot
>> be put back in. To validate the emotions, I think it is fine to be angry
>> about LOT=true and not like it, but we should either accept that it is most
>> likely to create consensus OR we should find a new game theoretic
>> activation strategy with better pro-social equilibriums.
>>
>> Personally, I think with either plan the ultimate risk of forking is low
>> given probability to activate before timeout, so we should just pick
>> something and move on, accepting that we aren't setting a precedent by
>> which all future forks should abide. Given my understanding of the
>> tradeoffs, I believe that the safest choice is LOT=true, but I wouldn't
>> move to hold back a plan of LOT=false (but would probably take mitigative
>> steps on community advocacy if it looks like there is non majority but non
>> negligible LOT=true uptake).
>>
>> Cheers,
>>
>> Jeremy
>>
>>
>> _______________________________________________
>> bitcoin-dev mailing list
>> bitcoin-dev at lists.linuxfoundation.org
>> https://lists.linuxfoundation.org/mailman/listinfo/bitcoin-dev
>>
> _______________________________________________
> bitcoin-dev mailing list
> bitcoin-dev at lists.linuxfoundation.org
> https://lists.linuxfoundation.org/mailman/listinfo/bitcoin-dev
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.linuxfoundation.org/pipermail/bitcoin-dev/attachments/20210223/90c7bfe1/attachment-0001.html>

2023-06-07 18:29:01

- reply