[{"author": "Phillip Hallam-Baker", "text": "

Bitcoin Delenda Est

", "time": "2022-07-26T19:01:50Z"}, {"author": "Martin Thomson", "text": "

@Phillip Hallam-Baker this isn't about that. But it might be about NFTs. I don't know just yet.

", "time": "2022-07-26T19:08:45Z"}, {"author": "Pete Resnick", "text": "

Scanning through the document, I'm not seeing a gap analysis. That's the big question for me.

", "time": "2022-07-26T19:15:06Z"}, {"author": "Benjamin Kaduk", "text": "

Is Ekr just too close to the room mic, or is the mic gain too hot?

", "time": "2022-07-26T19:20:32Z"}, {"author": "Melinda Shore", "text": "

I don't worry that much about gap analyses, myself. I have other concerns, but those can wait)

", "time": "2022-07-26T19:20:41Z"}, {"author": "Pete Resnick", "text": "

@Benjamin Kaduk Mic is just very sensitive

", "time": "2022-07-26T19:21:13Z"}, {"author": "Shigeya Suzuki", "text": "

Maybe mic gain is too hot.

", "time": "2022-07-26T19:21:32Z"}, {"author": "Melinda Shore", "text": "

For the moment I'm assuming that the gateway is \"trusted\" in the context of the protocol running on that network, given that it's running some sort of consensus protocol.

", "time": "2022-07-26T19:21:46Z"}, {"author": "Pete Resnick", "text": "

He's not that close or being that loud.

", "time": "2022-07-26T19:21:50Z"}, {"author": "Dave Thaler", "text": "

If you define things as \"legally recognized\" then you need to take assumptions from laws (which vary by jurisduction...), so maybe you don't quite mean \"legally recognized\"

", "time": "2022-07-26T19:23:03Z"}, {"author": "Alissa Cooper", "text": "

What is being gained by purposefully not specifying what an asset can be?

", "time": "2022-07-26T19:24:13Z"}, {"author": "Melinda Shore", "text": "

Avoiding the \"b\" word, I think.

", "time": "2022-07-26T19:24:36Z"}, {"author": "Pete Resnick", "text": "

The slide that had the colored area indicates that only the distance between G1 and G2 (inclusive) was part of the protocol. That says to me that the spec is only talking about assets that G1 already has access to (i.e., that for purposes of the protocol G1 is trusted).

", "time": "2022-07-26T19:31:17Z"}, {"author": "Benjamin Kaduk", "text": "

That's fair. Do we have an existence proof that such a setup is possible?

", "time": "2022-07-26T19:31:50Z"}, {"author": "Dave Thaler", "text": "

I am not yet convinced of tractability if you really mean legally recognized but the two sides disagree on legal recognition (e.g., because legal recognition may require involuntary transfer, for instance)

", "time": "2022-07-26T19:31:59Z"}, {"author": "Martin Thomson", "text": "

What I heard here was that the means of identifying assets was not yet standardized, which surely has to be a bit of a challenge.

", "time": "2022-07-26T19:32:04Z"}, {"author": "Melinda Shore", "text": "

Yes, it would have to be. I think if you look at some related work it tends to take place in the context of a consensus network, in which the gateway is a full participant.

", "time": "2022-07-26T19:32:43Z"}, {"author": "Martin Thomson", "text": "

The work that Greg Maxwell was doing on sub-chains in bitcoin involved a process that was effectively sequestration. But that isn't a transfer.

", "time": "2022-07-26T19:33:35Z"}, {"author": "Dave Thaler", "text": "

if you say voluntary transfer, and remove the requirement of \"legal recognition\" then I believe it is tractable and the proposed architecture would apply. but it may limit the use cases to those that don't have regulatory oversight, or where the entire system is under the same legal jurisdiction

", "time": "2022-07-26T19:34:38Z"}, {"author": "Melinda Shore", "text": "

There are already people working on environments in which there's a gateway, running different protocols on each side. There's a problem here, which is what this work addresses in part, is what to do between two different gateways

", "time": "2022-07-26T19:34:54Z"}, {"author": "Tadahiko Ito", "text": "

Are there case where any of A's asset can be legally recognized in B, but any of B's asset is not. so that transfer would be one-way. I was wondering that is inside of scope.

", "time": "2022-07-26T19:36:30Z"}, {"author": "Dave Thaler", "text": "

(and by same legal jurisdiction I include cases covered by a treaty between organizations, in which case the treaty agreement is the \"legal jurisdiction\" I refer to)

", "time": "2022-07-26T19:36:36Z"}, {"author": "Pete Resnick", "text": "

@Dave Thaler \"Voluntary\" by whom? If (in the diagram) G1 doesn't want to give G2 the asset, it says no. If you're asking whether G1 can involuntary taking an asset from the client, that's out of scope, at least AFAICT from what was shaded in the diagram.

", "time": "2022-07-26T19:37:46Z"}, {"author": "Pete Resnick", "text": "

(\"Out of scope\" meaning \"not something defined in the protocol\")

", "time": "2022-07-26T19:38:08Z"}, {"author": "Martin Thomson", "text": "

@Pete Resnick how can you disconnect one from the other? I mean, you can say that it is out of scope, but the structure of the network might make that impossible.

", "time": "2022-07-26T19:38:39Z"}, {"author": "Martin Thomson", "text": "

I can't read these slides. At all.

", "time": "2022-07-26T19:38:54Z"}, {"author": "Dave Thaler", "text": "

@Pete Resnick that's why I asked the question, because the terminology slide implied in scope but the solution slide would only seem to work if it's out of scope. It was a problem statement presentation, so it was debatable whether the gateway solution diagram was in scope for a problem statement presentation :)

", "time": "2022-07-26T19:39:04Z"}, {"author": "Pete Resnick", "text": "

@Martin Thomson Might make what impossible?

", "time": "2022-07-26T19:40:01Z"}, {"author": "Pete Resnick", "text": "

@Dave Thaler I agree, the terminology slide and even the diagram have the clients as entities in the problem statement, but then the slide with the shading indicates that they are not part of the protocol. Confusing. But I suspect he actually meant that the shaded area really was the entire scope.

", "time": "2022-07-26T19:41:41Z"}, {"author": "Martin Thomson", "text": "

@Pete Resnick So if the goal is to end existence of an asset in one network and start the existence of the asset in another, then you might need to engage with the processes within both networks in order to have the transfer take effect. Just saying that the gateways handle that might be possible in some cases, but in others ruling that out of scope would just be magical thinking.

", "time": "2022-07-26T19:42:13Z"}, {"author": "Dave Thaler", "text": "

@Pete Resnick but a related question solution-space wise is can G2 involuntarily take asset from G1, if G2 has a legal recognition of being able to do so (e.g., G2 is the government and G1 is operated by a private entity that has violated laws). But I don't want to debate the solution space architecture, I want to focus on the problem statement, and then derive whether the solution is appropriate, not define the problem around a presumed solution per se.

", "time": "2022-07-26T19:42:19Z"}, {"author": "Benjamin Kaduk", "text": "

So do we need to solidify what requirements we have on the two networks to be able to support this protocol?

", "time": "2022-07-26T19:42:49Z"}, {"author": "Dave Thaler", "text": "

@Benjamin Kaduk if we consider the \"protocol\" fixed, then sure one could narrow the use cases to where it could apply. My preference is to not do that, but rather accept a problem statement (perhaps narrowed to what is tractable and would be used), and then update the protocol to match it as needed

", "time": "2022-07-26T19:44:57Z"}, {"author": "Benjamin Kaduk", "text": "

That would be my preference too, but we can always make some restrictions so as to limit to what we think is possible to actually achieve.

", "time": "2022-07-26T19:45:34Z"}, {"author": "Dave Thaler", "text": "

@Benjamin Kaduk it gets a lot more tractable engineering-wise if you remove the assumption that it has to be tied to legal recognition. Laws of governments and laws of physics are not necessarily in sync :)

", "time": "2022-07-26T19:47:49Z"}, {"author": "Benjamin Kaduk", "text": "

:+1: In this jurisdiction, the value of pi is 3.0 :)

", "time": "2022-07-26T19:48:21Z"}, {"author": "Dave Thaler", "text": "

:+1:

", "time": "2022-07-26T19:48:32Z"}, {"author": "Phillip Hallam-Baker", "text": "

https://www.bolero.net/company-overview/

", "time": "2022-07-26T19:49:33Z"}, {"author": "Pete Resnick", "text": "

@Martin Thomson I guess it never occurred to me that it was even possible to do something in protocol that would represent ending existence if the players in a given network didn't already agree with the gateway.

", "time": "2022-07-26T19:49:57Z"}, {"author": "Benjamin Kaduk", "text": "

As the current live discussion is implying, we might need some ability from the network to support an object \"disappearing\" or being \"transferred out\" of the source network.

", "time": "2022-07-26T19:50:28Z"}, {"author": "Pete Resnick", "text": "

I guess at that level, I do think the protocol will be either boring or magic.

", "time": "2022-07-26T19:52:09Z"}, {"author": "Phillip Hallam-Baker", "text": "

If people are going to raise a use case as an example, I think they need to research that use case and be aware of existing work in that use case. There are 200+ carriers who already use digital bills of lading.

", "time": "2022-07-26T19:52:28Z"}, {"author": "Martin Thomson", "text": "

I'd just like to point out that this is not a typical URI structure, where there is an authority involved.

", "time": "2022-07-26T19:53:52Z"}, {"author": "Martin Thomson", "text": "

@Wes Hardaker why can't @Alexandru Chiriac present the slides?

", "time": "2022-07-26T19:55:33Z"}, {"author": "Wes Hardaker", "text": "

@martin: knowledge only

", "time": "2022-07-26T19:56:12Z"}, {"author": "Martin Thomson", "text": "

@Wes Hardaker you should be able to transfer control. Maybe we'll have fewer interruptions that way :)

", "time": "2022-07-26T19:57:10Z"}, {"author": "Wes Hardaker", "text": "

@martin: yes, lets not argue about the formatting of the identifier today -- leave that to the wg

", "time": "2022-07-26T19:57:14Z"}, {"author": "Benjamin Kaduk", "text": "

The \"lock status\" of an asset. Sounds like we assume a network has an ability to mark an asset as \"locked\"...

", "time": "2022-07-26T19:57:31Z"}, {"author": "Pete Resnick", "text": "

So it sounds like I can only send an asset. I can't request an asset.

", "time": "2022-07-26T19:59:13Z"}, {"author": "Pete Resnick", "text": "

@Benjamin Kaduk From the perspective of the protocol, can't I just lie? (I.e., can't I just say, \"Yeah, locked it\"?)

", "time": "2022-07-26T20:00:20Z"}, {"author": "Benjamin Kaduk", "text": "

Yes!

", "time": "2022-07-26T20:00:30Z"}, {"author": "Martin Thomson", "text": "

Well, this is the point about not engaging with the processes within the network.

", "time": "2022-07-26T20:01:25Z"}, {"author": "Phillip Hallam-Baker", "text": "

Protocol seems a bit longwinded. I think I can reduce it substantially.

\n

IPR Disclosure: I reserve the right to patent and refuse to license for use with Proof of Work systems.

", "time": "2022-07-26T20:02:22Z"}, {"author": "Melinda Shore", "text": "

Again, I think some of these properties make more sense if considered in the context of consensus networks (that is to say, blockchains).

", "time": "2022-07-26T20:02:29Z"}, {"author": "Melinda Shore", "text": "

That said, my experience with pretty much anything to do with distributing computing in the IETF has been somewhere between unpleasant and dire.

", "time": "2022-07-26T20:03:00Z"}, {"author": "Melinda Shore", "text": "

distributed computing, that is

", "time": "2022-07-26T20:03:16Z"}, {"author": "Martin Thomson", "text": "

@Melinda Shore distributed computing or \"permissionless consensus\"?

", "time": "2022-07-26T20:04:33Z"}, {"author": "Melinda Shore", "text": "

Both! Which is part of the problem with doing this work here. The likelihood that the commitment subprotocol would receive enough expert review to guarantee correctness hovers somewhere around, well, a very small number.

", "time": "2022-07-26T20:06:16Z"}, {"author": "Pete Resnick", "text": "

Seem like we are moving more and more toward \"boring\".

", "time": "2022-07-26T20:12:45Z"}, {"author": "Martin Thomson", "text": "

@Bernie Hoeneisen you just faded out a little, that's all.

", "time": "2022-07-26T20:13:19Z"}, {"author": "Martin Thomson", "text": "

@Bernie Hoeneisen your mic was worse the second time.

", "time": "2022-07-26T20:14:54Z"}, {"author": "Melinda Shore", "text": "

Bingo.

", "time": "2022-07-26T20:15:13Z"}, {"author": "Roman Danyliw", "text": "

What is the MVP and how is it validated by the intended end users?

", "time": "2022-07-26T20:16:03Z"}, {"author": "Bernie Hoeneisen", "text": "

q: the 1st presenter mentioned the scope to be only GW-GW. Is it meant that the protocol will be designed to work only GW-GW or will there be room for a possible later extension to make this work end2end?This is relevant e.g. for cases where GWs are hacked.

", "time": "2022-07-26T20:17:08Z"}, {"author": "Melinda Shore", "text": "

When we fired up dinrg, we thought that having the RG would motivate people from that community to start participating in the I{R,E}TF, but that did not happen.

", "time": "2022-07-26T20:17:31Z"}, {"author": "Bernie Hoeneisen", "text": "

with end2end I mean system in network A to system in network B (via GWs)

", "time": "2022-07-26T20:17:55Z"}, {"author": "Martin Thomson", "text": "

I think that the question of e2e here is an interesting one. Because it is not at all clear what the endpoints are in a system like this.

", "time": "2022-07-26T20:18:02Z"}, {"author": "Melinda Shore", "text": "

Right, and it appeared in the documents that the endpoints are the gateways

", "time": "2022-07-26T20:18:43Z"}, {"author": "Martin Thomson", "text": "

This idea of proving the existence of a lock implies to me that a gateway is not the true endpoint. That means that a claim otherwise is hard to believe.

", "time": "2022-07-26T20:19:24Z"}, {"author": "Melinda Shore", "text": "

Good point.

", "time": "2022-07-26T20:19:46Z"}, {"author": "Pete Resnick", "text": "

GW2GW is boring. Allowing extensibility for e2e seems like it might be magical.

", "time": "2022-07-26T20:20:07Z"}, {"author": "Benjamin Kaduk", "text": "

Maybe MT should get in the mic queue to say that

", "time": "2022-07-26T20:20:22Z"}, {"author": "Melinda Shore", "text": "

This may be somewhat fuzzy, actually, if you go back to trustless consensus networks

", "time": "2022-07-26T20:20:35Z"}, {"author": "Melinda Shore", "text": "

Because in that case, the endpoint is the network itself, which is treated as a single entity.

", "time": "2022-07-26T20:21:01Z"}, {"author": "Kohei Isobe", "text": "

Internet Ecology Task Force

", "time": "2022-07-26T20:21:40Z"}, {"author": "Henk Birkholz", "text": "

green is the new security

", "time": "2022-07-26T20:22:19Z"}, {"author": "Henk Birkholz", "text": "

Green Considerations

", "time": "2022-07-26T20:22:29Z"}, {"author": "Denis Avrilionis", "text": "

The essence of SAT is to implement ACID using 3PC across heterogeneous systems to transfer assets. The only ACID implementation of such kind of protocols are for the moment implemented in cloud/on premise architectures. SAT is a generalisation of existing implementations

", "time": "2022-07-26T20:22:47Z"}, {"author": "Roman Danyliw", "text": "

If SAT is ACID using 3PC, what are the security properties to address?

", "time": "2022-07-26T20:24:25Z"}, {"author": "Denis Avrilionis", "text": "

All of the existing ACID plus decentralised cryptographic guarantees in case the underlying network being DLTs

", "time": "2022-07-26T20:27:17Z"}, {"author": "Martin Thomson", "text": "

How is this likely to change the fraud/theft dynamics?

", "time": "2022-07-26T20:29:40Z"}, {"author": "Martin Thomson", "text": "

...if it doesn't engage with the operation of the networks and only deals with interchange.

", "time": "2022-07-26T20:29:59Z"}, {"author": "Roman Danyliw", "text": "

Denis Avrilionis said:

\n
\n

All of the existing ACID plus decentralised cryptographic guarantees in case the underlying network being DLTs

\n
\n

I follow that in the end-to-end sense, but I thought that the protocol was agnostic to network behind the gateway.

", "time": "2022-07-26T20:31:47Z"}, {"author": "Tadahiko Ito", "text": "

some of \"protocol\" in today's discussion does not seems to be (automated internet communication) protocol for me. I feel like they should define protocol also.

", "time": "2022-07-26T20:32:08Z"}, {"author": "Martin Thomson", "text": "

not 100% sure that the IETF is the right place? I'm closer to 0%.

", "time": "2022-07-26T20:33:43Z"}, {"author": "Melinda Shore", "text": "

I really don't think it belongs in the IETF, independent of the question of the merit of the work.

", "time": "2022-07-26T20:35:17Z"}, {"author": "Martin Thomson", "text": "

If this room had 400 people in it, including folks from a wide range of institutions, I'd be happier, but most of the people here are not in the relevant business.

", "time": "2022-07-26T20:36:38Z"}, {"author": "Behcet Sarikaya", "text": "

ACID atomicity, consistency, isolation, and durability with three phase commit is what we are talking about?

", "time": "2022-07-26T20:37:02Z"}, {"author": "Pete Resnick", "text": "

Did he just blame X.509 on us?

", "time": "2022-07-26T20:39:49Z"}, {"author": "Martin Thomson", "text": "

A lot of things are our fault, but X.500 is only our fault to the extent that we chose to use it.

", "time": "2022-07-26T20:40:43Z"}, {"author": "Melinda Shore", "text": "

I am not anti-blockchain or anti-distributed computing (to the contrary!) and I do not think this work belongs here

", "time": "2022-07-26T20:45:18Z"}, {"author": "Melinda Shore", "text": "

(and I acknowledge that \"I am not anti-distributed computing\" may be one of the dumbest things I've ever typed).

", "time": "2022-07-26T20:46:15Z"}, {"author": "Melinda Shore", "text": "

Does that mean that the commitment subprotocol is out of scope?

", "time": "2022-07-26T20:46:58Z"}, {"author": "Richard Barnes", "text": "

lawyers don't need crypto

", "time": "2022-07-26T20:50:02Z"}, {"author": "Martin Thomson", "text": "

@Melinda Shore do you want that relayed to the mic?

", "time": "2022-07-26T20:50:17Z"}, {"author": "Pete Resnick", "text": "

\"legal mechanism\" == \"out-of-band mechanism\"

", "time": "2022-07-26T20:50:52Z"}, {"author": "Martin Thomson", "text": "

The idea that legal might be synonymous with out-of-band is bizarre to me.

", "time": "2022-07-26T20:51:36Z"}, {"author": "Martin Thomson", "text": "

That is, building a system for the exchange of value that does not include provisions for governance, etc... is out of this world.

", "time": "2022-07-26T20:52:35Z"}, {"author": "Pete Resnick", "text": "

@Martin Thomson You're right; it's not \"==\"; it's \"\u2192\"

", "time": "2022-07-26T20:52:35Z"}, {"author": "Yoav Nir", "text": "

Tough to answer if 3/4 of us don't understand the problem

", "time": "2022-07-26T20:55:30Z"}, {"author": "Martin Thomson", "text": "

Can we ask about scope? Because I'm almost 100% convinced that the proposed scope is inadequate.

", "time": "2022-07-26T20:55:31Z"}, {"author": "Melinda Shore", "text": "

Did more people raise their hands in response to the \"right standards body\" question than in response to the \"do we understand the problem\" question?

", "time": "2022-07-26T20:55:50Z"}, {"author": "Benjamin Kaduk", "text": "

More affirmatives, yes

", "time": "2022-07-26T20:56:05Z"}, {"author": "Benjamin Kaduk", "text": "

fewer total votes

", "time": "2022-07-26T20:56:11Z"}, {"author": "Martin Thomson", "text": "

@Melinda Shore I can't explain it either.

", "time": "2022-07-26T20:56:18Z"}, {"author": "Benjamin Kaduk", "text": "

The scope of the proposed charter ... is not.

", "time": "2022-07-26T20:56:48Z"}]