- From: CCG Minutes Bot <minutes@w3c-ccg.org>
- Date: Wed, 14 Jun 2023 18:23:05 +0000
Thanks to Our Robot Overlords for scribing this week! The transcript for the call is now available here: https://w3c-ccg.github.io/meetings/2023-06-13/ Full text of the discussion follows for W3C archival purposes. Audio of the meeting is available at the following location: https://w3c-ccg.github.io/meetings/2023-06-13/audio.ogg ---------------------------------------------------------------- W3C CCG Weekly Teleconference Transcript for 2023-06-13 Agenda: https://www.w3.org/Search/Mail/Public/advanced_search?hdr-1-name=subject&hdr-1-query=%5BAGENDA&period_month=Jun&period_year=2023&index-grp=Public__FULL&index-type=t&type-index=public-credentials&resultsperpage=20&sortby=date Organizer: Mike Prorock, Kimberly Linson, Harrison Tang Scribe: Our Robot Overlords Present: Paul Dietrich GS1, Mike Prorock, Dmitri Zagidulin, David I. Lehn, Brian Richter, Joe Andrieu, Hiroyuki Sano, Japan, Geun-Hyung, Will, Bob Wyman, Jeff O - HumanOS, TallTed // Ted Thibodeau (he/him) (OpenLinkSw.com), Erica Connell, Harrison Tang, Greg Bernstein, Keith Kowal, Nate Otto, Nis Jespersen , Leo, Andres Uribe, Chandi Cumaranatunge, Kaliya Young, Kerri Lemoie, BrentZ, John Kuo, Orie Steele Our Robot Overlords are scribing. Mike Prorock: Her son didn't make it. Mike Prorock: https://lists.w3.org/Archives/Public/public-credentials/2023Jun/0037.html Mike Prorock: Hello all and welcome to the weekly ccg meeting for June 13th this is the year 2023 last I checked the topic for today is posted to the list is the Bitcoin ordinals did method which should be an interesting topic to discuss I think there's some great technical Merit here and I am. Mike Prorock: Really curious to see. Mike Prorock: https://www.w3.org/Consortium/cepc/ Mike Prorock: The walkthrough of one how we got here and to how it works so definitely interested in it with that I do want to start off the meeting by reminding everyone that there is a code of ethics and professional conduct and I obviously this as well as all w3c meetings are conducted under that code of ethics and professional conduct so hopefully we won't need any reminders for that during the call typically don't but just have to bring it up. Mike Prorock: https://www.w3.org/community/credentials/join Mike Prorock: Intellectual property note anyone May participate in these calls this is a public call however any and all substantive contributors to actual work items must be members of the ccg agree so the w3c credentials community group with an IPR agreement signed and I'll put the link to join if you're for some reason not a member the you will need a W3 account for that but that's free Etc and it can walk you through that so as. Mike Prorock: As with all of our meetings. Mike Prorock: And meetings will go up to our GitHub along with the audio recordings cetera we do manage chat via the jitsi side which is also linked to IRC in case you're in Via IRC but if you want to type it in the letter q followed by the plus sign will add you to the queue the minus sign after the letter Q will subtract you from the queue or you can click raised hand in that UNC interface and that will also put you on the Queue so with that I want to pause. Mike Prorock: As quickly any introductions reintroductions as anyone changed. Mike Prorock: Job new role or as anyone brand-new to this meeting that would like to introduce themselves Paul's here. Mike Prorock: Seeing no one come off of mute I'm going to also just side note that the transcriber is freaking terrible today with or especially with me so anyways with that I'm going to move on any announcements I know I have one for sure just for folks to be aware of but anyone else have any announcements for the ccg to me. Mike Prorock: Watching Q here. Mike Prorock: Not seeing any I am going to note that if you are a member of the VC working group you are already aware of this but if you're not but do monitor the list cetera a lot of stuff is starting to work its way into horizontal review so this is privacy internationalisation and tag etcetera so there could be changes coming to some of that based on feedback from horizontal view so just be watching that stuff closely if you're implementing on that sign. Mike Prorock: With that I am going to move. Mike Prorock: https://github.com/ordinalsreserve/btco/blob/main/spec.md Mike Prorock: To the main agenda here so we're fortunate enough to have Brian Richter here to talk today about the Bitcoin ordinals did method so basically it did method that mixes the security idea of blockchains particularly the Bitcoin blockchain with ordinal Theory so there is a spec for this and I'm going to copy and paste that directly into the chat and then I'm going to just hand the ball to Brian to take it away and I will be monitoring. Mike Prorock: During Cube to jump in and out. Mike Prorock: You know put folks on the point to ask questions at appropriate pause stuff so with that Brian over to you. Brian Richter: Cool thanks Mike yeah so come on everyone my name is Brian Richter I have a company and I've been involved in this space for a few years now companies a very Tech some of you might know me some of you might not but yeah I'm excited today to present to you guys the Bitcoin ordinals did method the the spec that Mike link there is the latest currently. Brian Richter: But there's a lot of changes in this. Brian Richter: Haven't made it in there yet so looking forward to moving things forward so I'm going to share my screen here hopefully that'll work. Brian Richter: That can be true. Mike Prorock: You are coming through nice and clear. Brian Richter: Yeah so the did BT CE o-- the start I don't think this group really needs an introduction to decentralized identifiers everyone pretty much knows it I think but I want to point out the three properties in the middle there are extremely you know very pertinent to the Bitcoin so Bitcoin is very decentralized as we as we know it's persistent the hash power behind it as. Brian Richter: It makes things quite difficult to change and. Brian Richter: It's also quite easily resolvable there's tons and tons of Bitcoin nodes that have exist in the world and basically you can get the data for this did method from from any of those. Brian Richter: So let's move on to the meat of this which is ordinal inscriptions so what the heck is that that's a word I've heard before ordinal but I don't know how that relates to bitcoin basically so it's at the end of last year like maybe December so this guy named Casey Radha more kind of created this this Theory ordinal Theory which essentially gives a number an identifier to every Satoshi in Bitcoin. Brian Richter: So a one Bitcoin can be broken down to. Brian Richter: I forget how many but a. <mprorock> 100 million satoshis to one bitcoin Brian Richter: Satoshi so it's two issues are the smallest unit of account on bitcoin and essentially using ordinal Theory you can give it each of those satoshi's and number which the number can be translated to different formats like the decimal format which is the Bloc numbered dot the offset within the block there's names which just uses English characters and has a mapping that he's come up with and then there's a few other formats like a degree no. Brian Richter: Location different formats the one I've been. <tallted_//_ted_thibodeau_(he/him)_(openlinksw.com)> (as usual) Link to deck? Brian Richter: Most is the the name version which kind of gives you the possibly a human-readable things often it's just random characters but you can find hidden human-readable Satoshi names. <mprorock> @ted - will make sure that it gets posted to list Brian Richter: So these sets they are just held in regular Bitcoin wallets there's nothing special about them they are just in sitting and utx Os that you control with your private key and you can kind of look up what satoshi's are included in a UT x 0 using the ordinal client which is an index on top of the Bitcoin note that you would also have need to be right. Brian Richter: So you can also you can. Brian Richter: Is so she's there's like I said regular Satoshi so you can transfer them just like regular using regular Bitcoin transactions the only caveat to that is that you need to. Brian Richter: Basically is ordinals aware because because they are regular sets they can like easily if a ball while it doesn't know about them will just accidentally spend them as fees or you know if you're trying to pay someone it'll just it'll just play with them because it thinks that they're just regular sets. Brian Richter: Then the most important part to this all is that they can be inscribed with arbitrary data so this uses segregated witness Taproot data and on the right there you can see the envelope that you've essentially kind of put the data inside of so there's a push saying just ORD and then another push that tells you the content file type and then another push telling you the actual content so this can be just you know any any sort of. Brian Richter: Of content and also. Brian Richter: Any sort of life that you want. Brian Richter: That's what inscriptions are there yeah like that arbitrary data that's written to the Bitcoin Ledger and you know it can be pruned off but it's also exhibits on all the nodes by default kind of thing and these are they're immutable these data blobs are in a mood to be linked to that sets the ordinal that you're writing to you can basically always look up the data that was inscribed to it and what this does is it creates unique Bitcoins. Brian Richter: Native digital artifacts so. Brian Richter: Mark outlined what a digital artifact is and the most you know pertinent most important things are kind of like their complete they're not links to off chain data they are you can hold artifact and it holds all of the information that's needed about it and doing all this you get something that is as durable immutable secure and and decentralized as Bitcoin itself because it is just Bitcoin with some you know layer. Brian Richter: He's on top of it which are embedded in the network. Brian Richter: I'm so the current uses that people have come up with is they've been writing just regular images into these inscriptions a lot of people are just been copy-pasting other entities and uploading them onto the coin and saying this is an interesting new thing and you know obviously it's nothing special but that's what people are doing or buying them selling them and there's also a naming system like a lot of people come up with these really pretty dumb protocols. Brian Richter: Which is just the first. Brian Richter: I have a name onto the thing is one who owns that name you can transfer that whoever you want or whatever like that doesn't have any of the properties that we want and dids but that's the come another thing that's kind of popped up is the arc 20 which is a token protocol like a fungible token protocol on top of this it's not the best for call it's you know costs a lot to just be moving tokens around so it's not. Brian Richter: It's not nothing special but people doing it people. Brian Richter: These tokens and you could also just embed you know like a full HTML page on there so there's people uploaded like a cone of Doom and the guy some Minecraft lots of different things and then some cheeky people and me and my brother created a couple verify your credentials put some art that he did in there and called a pub Collectibles and we give those people there's also you know kind of unbounded future. Brian Richter: He's with this like it's just. Brian Richter: The other day that you can own and by itself so it's interesting to see what people are coming up with and obviously as I'm discussing here another important use case I think is putting dids inside of these inscriptions so how do we do that I'll go through all of the different operations in the did method so the first of course is the create operation I wanted to make sure that there was you know. Brian Richter: A few. Brian Richter: To do it that all kind of work the same basically I want you to be able to upload or inscribe just the fold did document that you want to use so you could just create a did document embed the sap that you're writing it to into the identifier as you can see on the right there that one hij is the name of that side so this did document is is for that it's out there. Brian Richter: I've also. Brian Richter: Want to come. Brian Richter: The compressed did document format so to save a little bit of transaction fees because this is quite a lot of data that you're writing it could potentially get quite costly so compressed format is it's pretty important I haven't you know haven't landed on exactly what to do there but there's a few different ideas like that one on the behind there is the kind of the it's basically did pier and I'm not go too. Brian Richter: What's a little bit of a prefix there to. Brian Richter: Point out that it's a did I've also thought about maybe using compression like Sea border or broadly or I'm not entirely sure yet but I think there's going to be a good standardized way to do that I also had a call with one of the ordinals devs and he was pointing out that you could also do basically key-value stores in here so that's another possibility just kind of. Brian Richter: Betting the keys of the deduct human. Brian Richter: Directly with their values as possible. Brian Richter: So that's how you create and then when you want to look one of these things up to resolve it basically you grab the method specific identifier and the top right there you can see L WM l dot Theta whatever you look up what you T XO that sat is currently sitting in you got that you T XO and you. Brian Richter: You basically look. Brian Richter: What the first sat in that you txo the first inscription that you find because if you've updated then the current state is is is very first sat in there and then you read that inscription you find the current state if that. Brian Richter: Is the the null inscription which I haven't really defined yet but basically you're saying I'm deactivating this this did I'm not going to use it right now then it's sitting there as in a deactivated State you could also burn this in the description by setting it to a well-known burn address and then no one be able to move it kind of thing. Brian Richter: And that's basically. Brian Richter: How you looking look up the current state. Brian Richter: As I alluded to before to update basically you inscribe a new document into a noose at and you transfer along with that the original GID so the one that people will be looking up has to be in sitting that you txo probably at the end of it. Brian Richter: And then you can optionally you can. Brian Richter: Would all the. Brian Richter: The states as well so you can basically build this package of all the states that have happened for your D ID you could also kind of compress those those States and you know take the value out of those those previously inscribe States because there's at the end day there's sat sitting in there so there's kind of money. Brian Richter: And then. Brian Richter: To deactivate like I said you can update it to the no inscription which is kind of saying I'm deactivating this right now but I might be able to turn it back on the future so that's reversible deactivation and then there's also the ability to irreversibly burn the inscription. Brian Richter: So this obviously has some advantages and disadvantages this method the big advantage that that I like about this and the reason I created or made it is that you know this is layer one Bitcoin this is there's an 0 part counterparty risks there's you know nothing else you need to look up everything is within your Bitcoin know basically of course you need to have the. Brian Richter: Mapping of sets to the. Brian Richter: UT EXO's but that's that can be derived from the Bitcoin node so it's always have a Bitcoin node you have everything you need basically. Brian Richter: And that I believe basically creates the biggest possibilities for Network effects in as far as decentralized Ledger's go because Bitcoin is the most the most well-known blockchain out there it's the first and it you know it's easiest to get it's the easiest to kind of sink because it's the most there's almost no weeds and most decentralized anything. Brian Richter: Of course that's has some. Brian Richter: Method so you do want to wait for one block confirmation I would say to kind of trust that it did document is written and is the current state so it's a bit slow of course you could just kind of read them from the mempool if you wanted as well and assume that it's going to be eventually inscribed but you know that's you have to lay your wrist calling it there the also because the fees rise as more people use the network with them. Brian Richter: The way that the current bitcoin protocol or Bitcoin Network. Brian Richter: X is like there. Brian Richter: The market basically so you kind of have to bid for your block space so not too long ago their the fees went up kind of a crazy amount it costs $16 to send a regular transaction and then these ones that have a bit more data in them so they cost would cost you more than that so at some point it's kind of not really reasonable to use this method but for the most part right now the fees are low enough that it it's possible that kind of get what you want in. Brian Richter: So those are the trade-offs I think you know for some use cases the advantages overcome the disadvantages definitely but it's kind of we'll see so what work is there left to do I would say like the compressed format I mentioned isn't quite ironed out you know the burn address and what a known scription is there's a few little details I need to kind of work out and then I want to do another pass of the the spec and make sure it's all. Brian Richter: Up to you. Brian Richter: And then obviously I have to build an implantation I haven't quite done that yet I've done a little bit of the first operation creating I've created some and put them on the Ledger but I haven't haven't kind of gone through the the rest of the operations but especially because the specification is not entirely ready I would say it's a bit more to kind of research and then once that's ready I'll make a universal resolver put this up on the you know allow. Brian Richter: Anyone to kind of. Brian Richter: The universe of resolver so you don't have to run a Bitcoin now because I don't expect the normal user to want to do that so. Brian Richter: And yeah so I'd say I'm hoping to get some some interest in moving this along I want to start doing weekly meetings starting on Thursday I think around 10 a.m. probably and then this this coming week I'm hoping to have the first one I've created a Google meet I don't really know the organization of where this will be but I want to start with Google meeting kind of talk to people that are interested in I hope you hope you are there hope you interested in coming so yeah that's kind of the. Brian Richter: Station for today I have a couple Advanced topics if you guys want to get into those or if there's any other thoughts or questions. Mike Prorock: Signal went on the Q yet I could start it off with kind of to two quick questions I guess one would be well start with one actually since to be true just queued up the one would be are you concerned at all about. Mike Prorock: You know like transferring someone's Identity or sale of identities etcetera and things like that. Brian Richter: Yeah so that's a great question I've thought about it a little bit I'm interested in hearing what kind of the consensus for that is I've kind of I've heard both sides of it like you should be able to kind of transfer these identities to a different private key or different ownership kind of thing and different person but obviously that has attack vectors of impersonation and all that so my original update mechanism. Brian Richter: Like what. Brian Richter: And a transfer was basically burning that d i d and I'm not against moving back to that I want to hear kind of what other people think but yeah essentially there's a way we could design this so that transfer would end up deciding deactivating the ID so you wouldn't be able to kind of sell these things. Dmitri Zagidulin: So my question is is there a storage limit per single transaction how does I'm not sure I fully understand the ordinal model being talked about how like what the cost per byte is basically War if there is a storage limit to the. Brian Richter: Yes there definitely is a storage method how it works is I think this push here with the like hello world if this is too long I think there's a limit to the amount that you can push in there so you can but you can all continue to break it up into multiple pushes however there is essentially a limit I think it's around like 400 K 3 and k or something of transaction itself or else it won't be kind of relayed by the Bitcoin Network. Brian Richter: So there is there's limitations on that regard. Brian Richter: You can't get too large of like you know if your did document was a megabyte you might you wouldn't be able to do that really but for for the most part it's a reasonable limit and then as far as the cost so this is a very extreme case it's roughly around actually like 20 to 30 these days which is. Brian Richter: I would say. Brian Richter: To get a compressed form at 22:30 set v b on there for around like 50 cents a little bit less but it could go up to like five or so dollars now interestingly so because this is in the the segregated witness data you actually get a only cost you 25 percent of the set v b pricing because he had a witness debt discount on that which is kind of funny. Brian Richter: Um yeah so I haven't thought too deeply about it but I like if you are in control of a SAT and you know the name of it or the the idea 10 to power of it you can you could definitely create a did document that looks like this or whatever and you can have the the SAT name in there and stuff. Brian Richter: Like inscribing it on to the the ordinal would obviously make it discoverable but you you would be able to I guess send this did document before you inscribe it and kind of show someone and then you could also prove ownership over that it's at as well so that's an interesting idea for sure you could. Brian Richter: You could you could definitely play around with that. Brian Richter: Totally yes oh so interestingly so the the way that the actual ordinal inscription works is there's there's actually two transactions there's a commit transaction which is basically like what like what you said it's your kind of putting the commitment on to The Ledger whatever and then there's the reveal transaction afterwards that reveals the data and actually embeds the did it data. Brian Richter: The you could just do the commit transaction and go around showing people your to document whatever and then at some point in the future describe it if you want that kind of thing. Mike Prorock: Watching for other questions here as far as standardization path I mean obviously you've got a good initial spec up on. Mike Prorock: No up on GitHub etcetera I mean are you do you have you figured out is this something you think you you know want to roll down the diff direction or ccg or something else or if you looked at Pros cons Etc. Brian Richter: Yes I've a little bit kind of you know thought about where to kind of do this work for sure the I do want to make sure that it's open to you know it's inviting to people that are in the kernel space that actually kind of know how this stuff works it might be kind of. Brian Richter: Need to kind of get those people to come to the ccg you and his group that they don't know anything about or if where it seems like the essential identity and they're more interested in ordinals and Bitcoin stuff so I'm you know trying to figure out where exactly the best best place to do that is open suggestions for sure. Mike Prorock: Cool and then I guess the other side is this something you think that as an approach could be more generalized and have that going as one potential aspect but then permit use of other chains is that something you've looked at you know because obviously Bitcoins we got some uniqueness as far as like having a set break apart I think it's 100 million or something to one from a you know ratio standpoint of SATs to you know be TC but. Mike Prorock: It seems like other chains even even private letters could. Mike Prorock: You have similar type properties right is that something you've looked at possibly exploding. Brian Richter: Yeah totally so really the method like I'm trying to making sure to make it kind of like an agnostic way so that as long as the the chain supports ordinals you can you can do it and like for example like there's a doge ordinals protocol there's Litecoin or nose protocol and all of them I don't see a reason why they wouldn't be able to support the exact same thing. Keith Kowal: Hey thanks brother looks great could you explain I didn't quite follow the update methodology so like in particular like sorry I'll just go to the verifiable crunch world like I am issued a verifiable credential with this did and then a year later there's an update operation so I certainly understand like in the updated sat it can reference the old SAT but if you were looking at that VC like what would your resolver do with your like I don't think the resolver can't look at all SATs and try to figure out the current state. Keith Kowal: Of the of the like how would that work maybe I'm just missing something. Brian Richter: Yeah so how I think it would work is so if your resolver like you know the date that you're trying to resolve the the D ID for and the resolver could basically look up at that date what was the latest block and from there the you can look you can query the join node for like the most recent you. Brian Richter: EXO that that's at was. Brian Richter: Took the second question was transferred in and find that basically would be like the current state right so from that block height that you're interested in which is the you can find from the most recent transfer of the setting question and then you can find the utx of there and find the first inscription some extent. Keith Kowal: Okay yeah I got it I thought you were creating an sorry yeah I thought maybe you lost the connection to the oil said okay thank you. Brian Richter: Now you have some like that's the the important thing is like the new state and the initial state are tied together in a UT XO. Brian Richter: So I do have a couple of advanced topics here which I think we did kind of touch on a little bit the pre rotation so the commit revealed transaction format of inscription is is definitely an interesting thing to play with being able to basically commit to a rotation of your of your did document I think is valuable if you do you know that your next key is going to be before you make that live you can kind of commit to it. Brian Richter: Trying to get to the same sort of idea that Kerry has ever lived and then the next kind of advanced topic is basically so because you're doing this update over time you're in each of those inscriptions has like an envelope of SATs in it it might be quite a bit of money that you've kind of accumulated into your you're basically your old States and you can add some point you can basically fuse those back and create money out of them and turn your kind of receipts into money and just kind of fun. Brian Richter: Importantly like all you really needed the end of the day is the. Brian Richter: I need a state and the initial state to tie those together all of the other states can kind of be looked at from the previous you know using looking at the blockchain and then yeah that's kind of the to advance things I wanted to mention. Mike Prorock: So one question I need to get a bit at transaction time the because obviously with transaction times being what they are along with costs this seems like it's something that's gonna you know not necessarily like a use and burn type ID and it feels like they're definitely use cases attached to this if you thought through some of the most valuable use cases for this from a user perspective where it may bring properties that aren't granted by other did methods. Brian Richter: So for me specifically like I'm really interested in using it not as like the as my like only identify right but as an identifier that kind of helps people discover my other identifiers because a lot of the did methods I've been using did key to peer and stuff like there is no discoverability for them so basically being able to discover using this big Network that exists Bit Coin and not having to worry about. Brian Richter: Any any other thing in the world is. Brian Richter: It's kind of how I want to be able to be found kind of thing. Brian Richter: But yeah so as far as like use cases and stuff I it's pretty open like it's you know it's a very. Brian Richter: It's yeah it's same as all the other did methods I would say okay. Harrison_Tang: They have questions in regards to the cost of earlier you mentioned about come in and rebuild you take when you come into the transactions or you pick and then you don't pay when you inscribe the data or how this kind of course works. Brian Richter: Yeah so you pay for both the commit is like not very much data I think it's essentially the hash that you're writing those is the only like big data piece that you're putting in the command so it's you do have to pay for both but the commit is not nearly as cost for costly but the the reveal is where you're actually embedding the data so that is where the cost kind of in a bit crazier. Harrison_Tang: David so the bigger the data the more costly becomes basically. Brian Richter: Totally yeah exactly. Harrison_Tang: Got it and how does the update cost works like when you update when you updated basically it's a commit and rebuild again basically right. Brian Richter: Yeah exactly so it's update is relatively the same price as a create I would say I haven't run the numbers or anything on this but really like what is costly in in inscriptions is the obviously inscribing itself transferring them after you inscribe them is is a regular Bitcoin transaction so there's not too much fees associated with that but of course you're if you're transferring a whole bunch of this deal previous States or whatever you're going to be it's going to be a bit more. Brian Richter: Or data that you're kind of moving so you got to pay a little bit more but yeah. Brian Richter: You're paying the same. Brian Richter: Around the same every time you update. Brian Richter: That is as long as the current fees are you know relatively stable with if they going through skyrocketing then the thing the current market rate kind of thing. Mike Prorock: We'll look again. Brian Richter: Hmm yeah I mean I definitely thought about it I'm kind of of the opinion like like that the inscription should sort of be self-contained obviously that is more costly but I think like kind of the ability to just read that one inscription to know the current state is pretty powerful versus kind of you know driving the state from all of the. Brian Richter: The updates or whatever from. Brian Richter: Definitely for Simplicity but then of course there's a trade-off it is more costly so I don't know laying those pros and cons might be a thing that why should you should explore a bit more that's good a great idea though. Mike Prorock: Watching for additional questions on the Queue here. Mike Prorock: Any additional kind of unique properties you think are worthy of highlighting or areas of either security concerns or things like that that you may want to Dish analyze and on from the community side. Brian Richter: Yeah that's a good question so. Brian Richter: Like I feel like it's the probably the same as other did methods in this regard but like if you lose the control of the private key that controls those SATs you are essentially kind of losing control of your your identifier right I think most death did methods kind of have this property right like that's that's how public private key encryption you know drug-free workplace. <orie> sometimes there is guardianship / custodial recovery Brian Richter: I am wondering if there's mechanisms that should be built into kind of being able to like signal like this did has been kind of control has been lost with this did I need to move to this one here that kind of thing yeah so are you mentions guardianship custodial recovery and yeah that's kind of I'm getting to is like. <orie> but then that is easily exploited, with legal processes. Brian Richter: There are things that we should I should think about building into the specification itself in that regard or should I kind of want that question. Mike Prorock: And I think I'd love to hear or he's thoughts on the process and sees chatting very nicely on the tech side because I know he's thought about this issue so far away Ori. Orie Steele: Yeah I mean I think it just briefly like the did specification doesn't distinguish between updates that are made by the did controller updates that are made by the did method operator under duress or whatever so like just be careful with the word recovery like ion and diff like made it look like the recovery was a did operation. <harrison_tang> Christopher Allen's Blockchain Commons is doing the "Collaborative Seed Recovery", which tries to solve this type of issue Orie Steele: Update is a did Operation now if you decide that you want to delegate update capability to someone you really trust like your lawyer or your government or your priest that's your call and if that's what there's a way to cryptographically verify that update operations coming from another party that you delegated to previously that's a valid thing for you to describe in your particular did method but it's did method specific and it's a variation on the update. Orie Steele: Parameter and. Orie Steele: You can build your did method to expose that capability or you can choose not to expose that capability because now people can be coerced into giving those rights away it's sort of like the power of attorney thing like you know you hear cases of people being convinced to sign power of attorney documents when they weren't really sound of mind and then you know later that causes the problems for them so you'll be super careful with delegating to parties. Orie Steele: And making sure that that delegation. Orie Steele: A way that's safe and of course like once that delegation is occurred there's now two points of attack for the attacker they can go for review the did controller or they can go for any of your Guardians and use that as a way of sort of needing to get to you they can just go straight to your guardian and rotate keys. Brian Richter: I think there's definitely things that need to be thought about that for this method method as well as all didn't did essentially. Brian Richter: Um yeah and then the other kind of question I have for the community which might be brought up earlier is is like this update method is this a good approach do you guys think should I move back to like sort of a methodology that kind of limits the ability to transfer these things or is this do you think. Dmitri Zagidulin: Sense yeah so my question is so what does it mean to transfer a did be Co said that you hand over keep on Burger essentially delegating a particular dress tunic you. <orie> is it like buying a house at a specific address? Brian Richter: It's like an actual Bitcoin transfer of the SATs that descriptions are written on so it's yeah it's basically like well one way would be to transfer the SATs to someone right to a different private key different address and then of course the other method would be to give them your address earlier so your private key which would be the same as any sort of did method. <orie> (lets not encourage private key transfers) Dmitri Zagidulin: In that regard I don't think you as the author should worry too much about trying to prevent right because at the end of the day people can always hand over their private keys right like that it's not possible to prevent the transfer of something like a did even in theory. <orie> (you can't stop it, but you don't need to recommend it) Brian Richter: Actually and yeah or he's putting out in the comments like let's not encourage private key transfers obviously we don't want that but yeah like I kind of agreed to meet re like at the end of the day like being able to transfer this did document it did state to yourself at a different gear whatever should be evaluates for sure so yeah. Brian Richter: To answer your previous question already it is sort of like buying a house has specific address like that's why the inscriptions kind of how they work so yeah. Brian Richter: I think Joe's out person again. Mike Prorock: Joe I think you're up. Joe Andrieu: Okay thanks yeah I want to pile on this plus one about not limiting transfers I think one of the key reasons is Bitcoin is a specific solution to a double spend problem and unless you're going to allow that you probably want a less expensive substrate like Bitcoin gives you that for a reason and you pay for it so I think you should allow it otherwise you're going to undermine why are we using Bitcoin at all but I think for. Joe Andrieu: Perhaps some of your concerns about. <orie> why are we using bitcoin at all? <orie> spicee Joe Andrieu: Rated if you think about the difference between identifiers and identity what you're selling is an identifier the identity is is what results from the use of that identifier and so you don't compromise an identity just because someone transfers and identifiers and I think if you think about your the security of your systems and that way you'll be better off. Brian Richter: Yeah that's that's good point thank you. Mike Prorock: Pizza think you're up. Keith Kowal: Maybe maybe I'm missing like I would have thought that it doesn't matter in this but it doesn't have to matter who is controlling the sad or who owns this at I would have thought that maybe just an update operations you would want the signature of the controller of the private key identified and they create operation or deactivate operations and like I think that's what's kind of interesting about this is it doesn't matter who whose wallet this at sits in. Keith Kowal: But maybe I'm missing something. Brian Richter: No I don't think you are so like the set or the sort of the key that owns the utx so that the status sitting in that the document is sitting in is not related to the did itself right like it's the owner its kind of the owner of the did but it's not the. Brian Richter: Embedded within it which was a you know relatively deliberate design decision kind of separating the owner keys from the did keys are was was what I was trying to go for there. <orie> amusing to see eddsa keys in a secp256k1 registry system. Keith Kowal: And I got designed like a method I mean theoretically a method operator could have a wall at world is sat sieve sit but then the controller of the SAT could actually direct update operations. Keith Kowal: Sensor that's not possible. Brian Richter: Yeah no that makes sense that's exactly right so there could even be you could even have services that are doing the kind of like the ownership side but you could be telling them what did Keys your keys you want and you did document and stuff right so those the separation of those concerns is kind of what I was going for. Brian Richter: Yeah he points out that seeing you need to 55 women keys in a sec P 256 K 1 is funny and I thought so too and that was the first thing that people commented on when I brought this up on Twitter was people were like well why didn't you use Bitcoin keys inside your kids and I was like well I don't have to so why would I. Mike Prorock: Awesome well with that I'm going to just watch the key for any last words questions comments. Mike Prorock: And I'll hand it over to Brian just for any closing remarks etcetera and just wanted to thank you for your time today and thank everyone for great engagement so. Brian Richter: Yeah thanks thanks for coming guys and thanks for listening like I said I'd love to hear your love to get more participation from outside of myself in this did method so I will send out the link to the school meet afterwards and hope to hope to see you guys there. <orie> good luck standardizing BTC Method at W3C! Mike Prorock: Thanks so much I'm going to stop recording now and please have a wonderful rest of your week. <harrison_tang> Thanks!! Mike Prorock: 102 Transcriber still running.
Received on Wednesday, 14 June 2023 18:23:05 UTC