[MINUTES] CCG Weekly 2025-07-29

W3C CCG Meeting Summary - 2025/07/29

*Topics Covered:*

   - *Administrative Items:* Ethics reminders, IP policy clarification,
   meeting transcription and recording distribution. Announcement of upcoming
   Internet Identity Workshop in October. Proposal to establish APAC-friendly
   CCG meetings.
   - *Verifiable Credentials & Cryptographic Event Logs:* Approval of these
   as work items. Discussion of upcoming Google presentation on ZK
   advancements at the Data Integrity CCG call.
   - *DID Method Standardization:* The primary focus of the meeting.
   Discussion included:
      - The large number of existing DID methods and the need for
      standardization.
      - The process of selecting DID methods for standardization, including
      established selection criteria (technical features, adoption,
scalability,
      sustainability, economic aspects).
      - The roles of the Decentralized Identity Foundation (DIF) DID
      Methods Working Group and the proposed W3C DID Methods Working Group in
      this process.
      - The concept of "DIF Recommended DID Methods," a process for
      evaluating and recommending DID methods independent of formal
      standardization.
      - Discussion of the Controlled Identifiers (CI) specification, its
      origins, and its relationship to DIDs.

*Key Points:*

   - There's a growing need to standardize DID methods due to the large
   number of existing, often non-interoperable methods.
   - A multi-organizational effort is underway involving DIF, CCG, and
   others to select and standardize DID methods. Selection criteria include
   technical features and adoption levels.
   - A proposed W3C DID Methods Working Group aims to formally standardize
   select DID methods, potentially including blockchain-based methods (a
   significant shift from earlier proposals).
   - The DIF is developing a parallel process to recommend DID methods
   ("DIF Recommended DID Methods") through a review process, independent of
   formal standardization efforts.
   - The Controlled Identifiers (CI) specification was explained,
   clarifying its history and its relationship to DIDs. It's viewed as a
   political compromise that aimed at preventing a community split.

Text: https://meet.w3c-ccg.org/archives/w3c-ccg-ccg-weekly-2025-07-29.md

Video: https://meet.w3c-ccg.org/archives/w3c-ccg-ccg-weekly-2025-07-29.mp4
*CCG Weekly - 2025/07/29 11:55 EDT - Transcript* *Attendees*

Alex Higuera, Chandima Cumaranatunge, Dave Lehn, Dmitri Zagidulin, Erica
Connell, Geun-Hyung Kim, Greg Bernstein, Harrison Tang, Hiroyuki Sano, Ivan
Dzheferov, James Chartrand, JeffO - HumanOS, Jennie Meier, Joe Andrieu,
Jonathan Bryce, Kaliya Identity Woman, Kayode Ezike, Mahmoud Alkhraishi,
Manu Sporny, Markus Sabadello, Markus Sabadello's Presentation, Otto Mora,
Parth Bhatt, Rashmi Siravara, Rob Padula, Ryan Grant, Ted Thibodeau Jr,
Vanessa Xu, Will Abramson
*Transcript*

Harrison Tang: Hey Tad and Mamu, it's always us…

Mahmoud Alkhraishi: Hello.

Harrison Tang: who are on time, which is good by the way. I love it.

Harrison Tang: Hey, Marcus.

Markus Sabadello: Hello. Yeah,…

Harrison Tang: Hey, Thanks for taking the time to jump on here.

Markus Sabadello: Thanks for the invitations.

Harrison Tang: So, we'll start in about three minutes. and then I think you
already know the drill. We'll just go through some administrative stuff
first and then we'll jump into your session right away.

Harrison Tang: So Ryan, you asked, do we have a text channel in Q? so we
use Google Meet. So the message is the text session, but it will not show
up in the transcriptions. And then if you have any questions or whatever
just use the Google Meets raise hand functionality and then I will moderate
the queue. All right. So let's start. welcome everyone to this week's W3C
CCG meeting. So today very excited to have Marcus here our old-timer Marcus
to talk about CD and did method standardization.

Harrison Tang: So C stands for controlled identifiers and des is
decentralized identifier. So before we get to the main agenda, I just want
to quickly go through some administrative stuff. So first of all, just a
quick reminder on to ethics and professional conduct. just want to make
sure we have respectful and constructive conversations that we always have
but I think it's always good to spend about 20 seconds to remind everybody
about that right next just a quick note on the intellectual property anyone
can participate in these calls however all substensive contributions to any
CCG guidance must be the member of the CCG with PR agreement signed it
should be pretty easy to do though if you have

Harrison Tang: questions in regards to the W3C community contributor
license agreement. please feel free to reach out any of the co-chairs.
these meetings are automatically transcribed and all the code call notes
video audio recordings will be automatically sent out in 24 hours. again
thanks to Manu's work. All right. just want to take a moment for
introductions and reintroductions. So if you're new to the community or you
haven't been active and want to engage feel free to just unmute or use the
raise hand function Google Meet. Brian, please.

Ryan Grant: Hi, I'm Ryan rant. I work with Digital Contract Design and
we're one of the participants in the DID BTC1 DID method. And I've been
around over the years, but I haven't been on recent CCG calls. So, just a
reintroduction. Thanks.
00:05:00

Harrison Tang: Thanks, we would love to have you here every Tuesday and
great to have you here as well. All right, anyone else introductions or
reintroduction? Don't be shy. we love people to raise hands and ask
questions and be more engaging. so I won't feel as lone All right.
Announcements and reminders. Any announcements reminders? Money, please.

Manu Sporny: Hey, just a reminder that we are having Abby and Matteo from
Google come into the data integrity CCG call this Friday to talk about some
interesting discoveries that they've made with the long fellow ZK stuff. so
if you have some time it is probably going to be a highly technical
discussion but if you're interested in privacy preserving cryptography and
things that our community can do here to kind of help in that space I think
please do come to the meeting on Friday.

Manu Sporny: I guess the other question I have for the chairs is the call
for adoption for the verifiable credentials over wireless and the
cryptographic event log stuff. I'm wondering if we have a determination on
that yet or what the next steps there are. that's it.

Harrison Tang: So yeah, so the cultures have discussed it and I think we're
all in support of it. So we will proceed to actually approve it formally.
do you approve it? Will mamu do you mind helping us approve it? GitHub.
Yeah.

Will Abramson: No, I will take a Sorry about it. I know that's been out
there. I think we just message on the group, Unless there's anyone in this
call who really objects. I've not heard any strong objections to adopting
these items as work items. So, I guess consider them adopted. I will
probably put a message on the GitHub issues and…

Will Abramson: then maybe I'll send out an email as well to the mail list.

Harrison Tang: Yep. agree.

Harrison Tang: So yeah, my personal feeling is that first of all CCG is a
place for incubation of new ideas. So we will welcome new ideas. I know
some people have questions in regards to having potentially duplicated work
but this is not a DID method standardization discussion, right? use to use
the analogy right so I actually welcome all new ideas and proposals so
personally hugely in support of it and I think all the cultures think the
same as well as well the community so we will proceed to actually approve
both proposals so yeah please

Kaliya Identity Woman: internet identity workshop number one coming up I
think of October in California and…

Kaliya Identity Woman: so if you want to be there yes

Harrison Tang: Sorry, Clea.

Harrison Tang: Sorry, Clea, your internet is in and out. So, we hear that
it's about internet identity workshop this October. Sorry, we lost you, All
right. I think KA is sharing that in there. the next internet identity
workshop is this October. I'll share the link here. you just share it.
Korea.

Harrison Tang: Anything else other than the IW in All right. I'll come back
to Kuya if she has anything else later in the meeting. Any other
announcements or…

Harrison Tang: reminders? Yes,…

Will Abramson: Yeah, I have one.

Will Abramson: Yeah,…

Harrison Tang: please. Will

Will Abramson: I mentioned this last week or a couple weeks ago and I've
had some great responses. So, I'll say it again and it's just a call for
people to reach out to me if you're interested. I hope to schedule I don't
know let's say a series of CCG meetings in APAC friendly time zone. So
probably in parallel to this call we'll do another set maybe in my morning
but this is a conversation that we need to have when would that happen in
October. So TAC's going to be in November in Japan. I think it'd be great
to use that as momentum and incentive to try and build up the community in
the AP pack region.
00:10:00

Will Abramson: And just from my last call, I think a lot of AP pack folks
don't get to show up to this call, but I think Harrison can speak to They
do listen to the recordings, and I said this last time and I've already had
two or three people reach out to me saying thanks and expressing support.
So it shows that there is kind of demand for this. so yeah, if you are
interested in this and want to help make it happen, do reach out to me and
I'm looking to maybe have a kickoff coordination call in the start of
September. and we'll just discuss when might this happen? day? and also
what content might we want to put on there? I have a few people that I'm in
conversation with who might be able to do APAC related talks. So, it should
be fun.

Will Abramson: It be nice to connect with a bunch of different people. So,
sorry people from the Pacific, you guys might not be able to come, but
that's tough.

Harrison Tang: Thanks Will.

Harrison Tang: I think it's a great idea. so thanks for taking the
initiative. and yes, I think our audience is a lot bigger than just people
showing up live in these meetings because I get these emails correcting me.
And by the way, I really appreciate it. Sometimes I will send the date
wrong or something like that so people do read these transcriptions that's
for and watch the recording. So I think our audience is a lot bigger than
just 40 people that show up regularly every Tuesday. So no great idea. So
thanks. Any other questions, announcements, reminders?

Harrison Tang: All right. Any questions on the work item related stuff.
last calls for reintroductions, announcements, reminders, and work item.
All right. let's get to the main agenda. So, this week, very excited to
have Marcus, who

Harrison Tang: let the ID methods standardization work to talk about
control identifiers and decentralized identifiers method standardization as
we know there's a lot of DID methods so when the implementers want to
implement it often times they have questions on which DID method to use so
I think this standardization work on the methods is quite important so
Marcus I think the floor is yours. and thanks for jumping on and to share
your work and…

Harrison Tang: the group's work as well. So, thanks.

Markus Sabadello: Okay, great.

Markus Sabadello: Thanks Harrison for the invitation and hello everyone.
So, yeah, I'm going to talk a bit about good methods or methods
standardization. I don't really have slides but I will show you some of the
resources some of the repositories and places where this is happening and u
tell you a bit about what's going on there. So as you said already Harrison
there are a lot of deep methods right.

Markus Sabadello: So in the W3C did working group there has been this list
of the methods for a while we've never really called it a official registry
or anything like that right it's more like a list of known bin methods in
the community but it's a very long list right and so often we hear
arguments that people don't know which

Markus Sabadello: which ones to implement. Sometimes we get some criticism
that these are not easily interoperable because there are so many DI
methods and they all very different and as we know there has been
standardization of bits themselves, right? So there's the 1.0 W3C
recommendation.

Markus Sabadello: there is controlled identifiers from the verifiable
credentials working group and now there is work happening on did version
1.1 to standardize that however there has never been a standardized D
method right only the core DI data model the D documents the D syntax that
is in the D recommendation
00:15:00

Markus Sabadello: But there's no standardized it method. I'll post some of
the links in the chat. So we said let's try to start a process to
standardize a few concrete D methods. We meaning a few organizations have
come together to set up a D methods working group. and this working group
this is hosted by div at the decentralized identity foundation but it's
really a collaboration between div and CCG here plus IP foundation and
inappata and a number of other organizations which support this work.

Markus Sabadello: So this is a working group in diff this is meeting on a
weekly basis right now on Wednesdays there's an agenda document with the
topics and recordings and I will share a bit what this group has been
doing. So this group at the moment it's not actually standardizing deep
methods yet because deep is not a traditional standards body like W3C or
ISO others. But the objective of this group is to help to mature and
advance certain bit methods and then to support the standardization process.

Markus Sabadello: when this group started to meet the first thing we worked
on was selection criteria, So we asked if we want to pick a few DI methods
and standardize them, how should we select the methods out of those 200
that exist already? And so we asked the community what should be a criteria
for picking a few bit methods out of 200 that should be standardized
initially. so here on this page you can see some of these criteria I'm and
they are from GitHub discussions and mailing list posts and discussions in
the working group meetings.

Markus Sabadello: I'm not going to read all of them. you can also see a
recording and some slides where some of these have been discussed. But
those criterias some of them have to do with technical features like
privacy preserving crypto or binding to DNS, some technical features but a
lot of the criteria also have to do with adoption.

Markus Sabadello: So is a dip method being used? how many implementations
are there? And then there are also some criteria that have to do with
scalability and performance daily transaction volumes or things like that.
And sustainability, for example, is it how also economic aspects, So, for
example, how much does it cost to create a deed, in a certain bid method.
so for example, if we selected let's say three or four DID methods that we
want to standardize, then there should also be some diversity with regard
to the selection criteria, right?

Markus Sabadello: So if we select let's say four did methods to standardize
then they should not all did methods where creation of a deed costs a lot
of money right or we should not pick for the methods which take some time
during the creation process. we probably want to pick at least one DID
method that is very efficient, very cheap and very fast. but we also want
DI methods that support a lot of features. Right? So the idea here is to
select a few D methods for standardization that together as a set of
methods support a lot of these desirable criteria.
00:20:00

Markus Sabadello: and this is obviously also related to some work that has
happened in the community before related to evaluating and characterizing
deep methods. So there's also a reference to the bit rubric which is a
framework for evaluating bit methods according to how privacy preserving
they are. And there's also reference to the deep traits work if you're not
familiar with that. D traits is about characterizing bit methods with
regard to their features. Right?

Markus Sabadello: for example, key rotation? Does a D method support
service endpoints? does a D method support the update operation and things
like that. also trademark and IP issues. and so this list of selection
criteria the idea is that this would help us find out which the methods
should be initially standardized.

Markus Sabadello: there has also been this idea what you see here at the
bottom for a while that probably it makes sense to standardize at least one
deep method from a set of criteria broad characteristics like ephemeral
deep methods like DT or the JWK web-based D methods like webs

Markus Sabadello: web plus fully decentralized peer d methods. So this has
also been part of the discussion to pick the methods from such broader
types if you want. so that's what we did for a while in this working group
and then the next thing we did was to actually ask for pos. So we asked the
working group and the community to actually propose concrete methods that
would be standardized. And so we came up with a template that we asked
sorry just copying and pasting Harrison.

Markus Sabadello: You raise your hand.

Harrison Tang: Yeah, just a quick question. So earlier in the selection
criteria, there's a bunch. what do you the group thinks are the most
critical criteria or should I say can you kind of roughly prioritize these
criteria to highlight some of the most important ones that the group thinks
is critical in these selections?

Markus Sabadello: I mean, we didn't want to spend too much time on this, we
could have said that we want to develop a comprehensive framework with
ratings and priorities and formulas to then do some calculations which the
method fulfills which criteria to what extent and then weigh that and add
it all up and then get some mathematical results. But we thought that would
take a really long time to do that. So of course there's some criteria that
are maybe more important than others. I would say maybe the ad the
implementation, but of course conformance with the specification and things
like that.

Markus Sabadello: But with regard to many of these criteria, it's more
about the diversity of DI methods that we select rather than which D method
on its own scores meets the highest priority criteria, we did this more of
as an exercise to understand the diversity, For example, just as an
example, if we said that support for key rotation has a really high
priority, then that's only true in some cases, What if you wanted a deep
method that is very efficient and you care more about that and you don't
actually care about evotation? So it's hard to place some objective
absolute priorities on most of those.
00:25:00

Markus Sabadello: It's more about picking a set of bin methods that meet a
lot of the criteria and…

Harrison Tang:

Markus Sabadello: then yeah so like I said then we started collecting
proposals for the methods and we created a template. So we asked the
representatives from the method communities to fill out here a few sections
to include a description of the D method and rationale why do you propose
it for standardization what are some existing materials what's the current
specification what implementations test suites and so on and then we asked
people how does your proposed

Markus Sabadello: method meet the selection criteria. So that's where this
relates to the previous topic of the criteria and then there also something
about is the D method already being standardized somewhere and what are use
cases maybe that are specifically important for a certain bit method. And
then if you look in this repository, this template has already been filled
out by quite a few bit methods. right. So we see some diversity the web
some ledger based methods. so I'm not going to go through all of them but
take a look at these and then

Markus Sabadello: you'll see that the community that has already proposed a
few different methods. what happened then was then we said what we want to
do next is basically we had a discussion of what should happen then with
these deep methods and how they should be standardized. Right?

Markus Sabadello: as I said earlier diff is and this working group a diff
is not really not a traditional standards body right so we can't really say
okay we'll pick a few bit methods here and just standardize them there is a
certain process in div for creating work items and then also publishing
work items but it's not the same as let's say a W3C standard or IETF
standard.

Markus Sabadello: So what happened then is that at the same time in
parallel as we are working on this deep methods working group indeed
there's also a process now and the proposal to start an official deep
methods working group in W3C right so it's a bit confusing because they're
both called the same thing they're both called the deep methods working
group But the idea is to create one in W3C. for that there's now a proposed
charter which I will also put in the chat and then if this is approved then
this group can actually push a few deep methods through the W3C
standardization process.

Markus Sabadello: and this charter some of this has been influenced or
advised by the work that we've been doing in this deep working group. One
thing that's very important to point out is that it's definitely not the
intention to standardize all the methods in W3C. Right? So the deep methods
concept will continue to be a concept that's decentralized right so there
is not one authoritative central place where all the methods have to be
standardized it will always be possible to have deep methods that will be
standardized elsewhere or not be standardized at all. one of one good
example is the EPSI method.
00:30:00

Markus Sabadello: EBSI a European blockchain service infrastructure that's
an EU project and they have a work item already in a European standards
body to standardize the deep HC method there right so it's not a
requirement that all the methods have to be standardized in a specific
place having said that this year this proposed methods working group at W3C
this is a place where if it gets approved, where some bit methods can be
standardized.

Markus Sabadello: And here in the description, we actually took some of the
types of DI methods or what I called the broad categories of DID methods
like an ephemeral D method, a web- based D method and a fully decentralized
D method. that's what this proposed charter says right now that the working
group will deliver the method specification or at least one for each one of
these three groups.

Markus Sabadello: it also lists some examples and these examples here are
linking back to the div deep methods working group this proposal template
that I mentioned. the working group can then still decide which did method
to standardize or a new one or did method other than those that are listed.
So this is not a requirement that one of those that are listed here will be
standardized in the group but it's a list of input documents right for the
group.

Markus Sabadello: we had quite a bit of discussion here about the out of
scope section. until recently this said that the methods which are based on
blockchain technologies are out of scope. So the Walton group here would
not be allowed to standardize blockchain or ledger based methods. was a lot
of discussion about that. we thought that there may be a lot of opposition
to a charter if it allowed blockchain based methods. This has been
historically a point of contention or a point of criticism. Whether it's
justified or not is another question.

Markus Sabadello: But historically some of the opposition against bits in
general had to do with blockchain technology. So we initially wrote here
that that would be out of scope. However, then the feedback from W3C was
that it's probably better to not mention any specific technologies here,
but instead to include a more general reference to web architecture or web
design principles. So yeah, personally I think that was a great improvement.

Markus Sabadello: so now if this charter is approved then there will be a
W3C did methods working group that can work on a few concrete did methods.
There's still a bit of an open question or some feedback that maybe there
should be different working groups for each DI method, So maybe one working
group should only standardize one did method and another working group
should standardize another did method but I think there's a lot of
agreement or a lot of support for this proposed charter as it is as it is
right now.

Markus Sabadello: okay so that's the W3CD methods working group. the diff
working group where we started to collect these proposals. Then what
happened next was that then we discussed what can we do now in this div
methods working group. if some DID methods will be standardized in W3C and
at least one D method will be standardized in a European standards body and
some other D methods may perhaps not be standardized in an official
standards body at all. which doesn't necessarily mean that it's a bad bid
method, right?
00:35:00

Markus Sabadello: So then we came up with a new idea which is what we call
div recommended did methods. So that's the latest thing that we are working
on now. here we thought it would be a good idea together with the other
organizations who are also collaborating here if we could define a
mechanism or a process where we can list what we call diff recommended did
methods.

Markus Sabadello: So that means did the methods that are not just listed
somewhere the 200 or more than 200 but did methods that have gone through a
certain review process right so we said that's maybe something useful that
the div group can do and it's a bit orthogonal to the standardization right
so we said standardization can

Markus Sabadello: happen in W3C or elsewhere or not at all. but it's still
useful then to have a place where as an adopter as an implement you could
go and you could see a list of the methods that have gone through a certain
evaluation and review process and maybe certain quality assurance. we're
not calling it certification, So it's not quite that. It's not as thorough
as that but we defined a few steps that a bit method can go through.

Markus Sabadello: So if you look at this page there is a big table and in
this table you can see a few did methods and those are again exactly the
methods that have filled out the template which I showed earlier the
proposal template and we said here are a number of things that would be
good for a bid method. to go through to then become more trustworthy maybe
to eventually reach a stage or achieve a label that we diff recommended. So
for example, it says is the D method in the W3C registry?

Markus Sabadello: has the proposal template been filled out with what I
showed earlier? Does it have a Or does it have an entry in the DI test
suite? Right. the W3C D working group has a test suite where did method
implementations can evaluate their D documents and their inputs and outputs
against the test suite. So does a DI method have a driver in the universal
resolver? has the did method filled out the did traits framework? Are there
multiple implementations? What kind of deployments are there? Where is it
going to be standardized?

Markus Sabadello: and then we've started to conduct a series of deep dives
in this working group where representatives of a deep method present the
method and give the group the opportunity to ask questions and review the
method. and so based on that we think if a deep method can show all these
things and has gone through at least two deep dive sessions in this working
group then it could achieve the status of diff recommended.

Markus Sabadello: at the moment no did method has that right so we've
started this process but I'm not finished it we've also spent a bit more
time documenting and discussing the process so that's also listed and
described here and still refining that a little bit but that's basically
what's
00:40:00

Markus Sabadello: what's happening now and again again this is a bit
orthogonal or complimentary to the standardization process right so we
think those are two different things some D methods will be standardized in
W3C or in another standards body and some D methods may not be standardized
at all but

Markus Sabadello: Independently of that, we also have been working on this
process which starts with a submission of a proposal. and then at some
point the method could receive this status. and yeah, that's the current
state of things. let me see…

Markus Sabadello: if there are any questions thoughts.

Harrison Tang: Great. Thanks,…

Harrison Tang: Marcus. So, if anyone have any questions great. Monu, please.

Manu Sporny: I guess the current question is what do you think the timeline
is Marcus for each initiative? I think we've got a charter that is
presumably going to go in front of the W3C membership sometime to soon.
excitingly with the possibility of blockchain based methods being in scope,
I think that's a great change. we have the did recommended things kind of
happening. I know that this work all of it initially got kicked off because
the European Union work on digital wallets and the fact that not having a
standardized DID method was one of the reasons that the European Union at
least some of the architecture documents decided to not put DIDs in scope.

Manu Sporny: I'm wondering if we're on track to address that criticism in
the European Union or it doesn't really feel like one of the or I'm
wondering if it's one of the main goals. I guess the general question is,
when do you see the first set of DID methods being on an official standards
track and when do you see the first set of diff recommended methods
happening? That's

Markus Sabadello: I'm good question. I'm not sure to be honest. I mean,
some of these things better maybe as we know the W3C standards process can
take a while. we probably need to update also this start date here and then
with an end date in probably too late for the EU implementing act for the
EU DI wallet to officially reference that.

Markus Sabadello: I don't think that's possible maybe but there is this
standards process in the European standards body SEN right sench is what
that is called that has been approved so there is an active work item in
the European standards party for not just did epsy but also for deeds in
general And that may be on time potentially for being referenced. I'm not
really sure to be honest. I don't know those timelines well enough. But
that's actually an active work item, right?

Markus Sabadello: So even before there's this W3C working group for DI
methods, there is already a work item with regular meetings in Europe for
did epsy and dits in general. And I don't know all I can say is that I
think all of this makes a lot of sense independently of that one European
UDI wallet project, right? you're correct that this is what triggered the
work here but it's something that I think will be useful for many more
projects also in Europe for example we have something else called the
digital product passport and there we have the same discussions should
digital product passports use dats not and which did methods and that has a
00:45:00

Markus Sabadello: timeline on the GTI wallet and there will be more
projects around the world. So I'm not really focusing my attention on just
that one initiative. but I think this is just what we've been missing in
the D community in general to add the standardization and here and answer
the question right which one of the 200 methods should be implemented.

Harrison Tang: Any other questions? Well played.

Will Abramson: I guess I have a question for the rest of the group, but can
you talk to us about the history of why the SID spec got developed and what
is the next steps for that specification?

Will Abramson: Yeah, it's controlled identifiers, right?

Markus Sabadello: Okay, I think yeah,…

Markus Sabadello: I think there are probably also other people who know
that a bit better, but I can say that after we finished the did
recommendation at some point then there were some ideas to generalize the
documentation

Markus Sabadello: data structure, right? And that's what this is.
Basically, controlled identifiers is a work item in the verifiable
credentials working group, not in the DID working group, but there was a
desire to use the DI document data structure with the service endpoints and
verification methods and those things to use that for identifiers other
than specifically HTTPS URLs. so for example, if you have a verifiable
credential and you want to identify the issuer of the verifiable credential
not with a pit but with a HTTPS URL, then you still need to discover the
public keys and all of that.

Markus Sabadello: And so the idea was that these HTTPS URLs can also return
or can be resolved to a data structure like the document and that's now
called the controlled identifier document which is in this specification
here controlled identifiers so it's a generalization of the D document but
it are compatible, So, it has all the same constructs, the same data model.
And now the next version of the did recommendation D 1.1 actually
references that, right? So, this does not anymore define the full it
document data model, but it references the control identifier specification.

Markus Sabadello: And basically it says that a t document is a controlled
identifier document where the identifier is a bit but there's not really
any breaking change in that. is just reorganizing the specifications a
little bit. so that the document data structure is now the controlled
identifier document is a generalized version of the deep document.

Harrison Tang: Man, please place.

Manu Sporny: Yes, plus one to what Marcus said. there's nothing that you
said that was inaccurate or that I disagree with Marcus. but I think also
Marcus is being kind of how that specification came to be. I have a
slightly different perspective. So I am the lead editor on the controlled
identifiers document. I did not do that work because I thought it was a
good idea. I did the work because there was a set of people that really
hated DIDs and they were going to fork the work unless we made this
compromise. So my view is that the controlled identifiers specification
came about under political duress and it exists for that reason.
00:50:00

Manu Sporny: from if we're looking at decentralization and trying to make
sure that we're building decentralized systems the controlled identifiers
document is not that it doesn't give you that right we could have did web f
folks that wanted controlled identifiers could have just did web it would
have given the exact same set of features but there was this insistence
that dids are bad take them out or we're going to fork market and that's
even worse for the ecosystem. So, we tried to take what we had in the DID
document and extract it. There was also stuff that was happening in data
integrity that some folks wanted to use. So, we tried to extract that stuff
and put it in the controlled identifiers document and layer things
appropriately from an architectural layering perspective. And I think we
were successful in doing that.

Manu Sporny: but fundamentally I view the controlled identifiers document
as a political compromise. that I don't think does an enormous amount of
harm because the very same people that wanted us to split that out then
ended up I don't think using the controlled identifiers document at all.
They just went on to kind of throw more shade at DIDs and not implement it.
So, as far as implementations for controlled identifiers go, I don't see I
don't know of anyone that's really really implemented the spec purely as an
HTTP sbased thing. I think most people are still using did web that kind of
wanted the same feature set and so on so forth.

Manu Sporny: and I really do appreciate Marcus the way that you framed it
but given this community I want to also recognize that there are a number
of people in this community that are pretty unhappy that we did the spec at
all and wanted to convey why kind of the motivation behind working on the
specification. it was to try and keep some semblance of the communities
together. but at the end of the day, I think the vast majority of us are
just focused on DIDs and we want to see DIDs succeed. and if folks want to
take controlled identifiers and make them succeed, that's great. But again,
as I said, I don't know of anyone that's actually, purely just implementing
controlled identifiers, everyone I know that's implementing this stuff or
implementing DIDS. that's it.

Harrison Tang: Thanks By the way, your explanation help me clarify a couple
things. I always have a question in regards to what's the context behind
did the CI split too. So great. Thanks. Any other questions? Ryan, please.

Ryan Grant: Yeah, I just wanted to add that as far as indelible records for
history, the existence of this is certainly going to make a mark that they
can't deny. There was deeply political motivations going on with the
specifications. So that's interesting.

Harrison Tang: All By the way, Marcus, I just have a small suggestion is
that because nowadays more and more people are using basically AI chat
bots, right, to do research. So, I think we can DID methods documentation
for what's called AEO. So technically it's called answer engine
optimization. Basically it's like a subd discipline of SEO. and so the
actionable suggestions I would make is that for example you probably want
to put the traits or those criteria in the table that you share earlier you
just say did traits.

Harrison Tang: Yes. Yes. Yes. Right. So, Encode just means that you want to
add more columns and then put the attributes there because if people ask
questions saying that, hey, I'm trying to implement DID methods and I want
to look for something that results really fast. That's my top most concern
which did methods we recommend? then you want to actually have the ideal
scenario by the way for AO is Q\&A format because that's how language
models doing the reinforcement learning with human feedback is trained.
00:55:00

Harrison Tang: So you want to do some kind of Q\&A format but I think that
might be too overboard here but at least encode the traits and attributes
in the table that will help AEO some people call AI engine optimization…

Harrison Tang: but technically it's called answer engine optimization there
are some articles about how to do it but that's just my suggestion. Yeah.
Yeah.

Markus Sabadello: Mhm. Yeah.

Markus Sabadello: First of all it would become a really huge table. al also
also…

Harrison Tang: Yeah. Yeah.

Markus Sabadello: what we want to do here with this idea of these
recommended methods we don't actually care so much what the trades are. I
think we would care that someone has taken the time to fill them out. So we
would not say for example that we're not going to not recommend that the
method if it doesn't support the rotation or we're only going to recommend
methods so that are very scalable.

Harrison Tang: Yeah.

Markus Sabadello: So it's more about the fact that someone take the time to
fill it out, right? It doesn't really matter so much what it is, but
there's going to be a link, we haven't done that yet, but yeah, just like
there's a link to the method proposal and there's to the results of the
test suite. there will also be then a link to the table in the traits work
item, right? So that the traits is a separate work item where there's a
table with all the actual trait values. So this will link to that other
table where the actual traits for the method are documented. So if I was an
AI bot, I would then compare these two tables, right?

Markus Sabadello: I would know as an AI bot, I would know that this table
tells me which did methods are recommended and the other table would tell
me…

Markus Sabadello: what are the D traits for a given method. And then the AI
bot should take both into account when answering the question. What the

Harrison Tang: Yeah, I guess my premise is that less and…

Harrison Tang: less people are going to read the original documents.
Rather, they're just going to ask these AI bugs and take the answer. Now
people might think that Hollimation is a problem but to be very frank most
people they just don't worry about that much right and take AI rack
retrieval augmented generated answers as the answer so you kind of want to
optimize for that just want to give some suggestions because I would
foresee more people using AI bots to get recommendations about what did the
method

Harrison Tang: is to use in whatever context they're in as opposed to
reading through this document firsthand. Right. So guess that's my guess
for the future.

Markus Sabadello: Yep.

Harrison Tang: Prediction for the future for that better term. All right.
by the way, one last thing. Kalia, we didn't really hear you other than
there's a IW coming up in October because earlier you had some internet
issues. did we miss anything else? Do you want to add last comments?

Kaliya Identity Woman: Yeah, I was saying that I just wanted to remind
folks that we're committed to accessibility. So, if people want to be there
and are challenged by our ticket prices to please reach out.

Harrison Tang: Thank you.

Markus Sabadello: Thank you.

Kaliya Identity Woman: Thank you.

Harrison Tang: Thanks, All right. we're at time. thanks thank you Marcus
for jumping on here and then lead a great discussion. this concludes this
week's CCG meeting. Have a good one. See you next week.
Meeting ended after 00:59:56 👋

*This editable transcript was computer generated and might contain errors.
People can also change the text after it was created.*