CACAOv3 #196

oed · 2022-12-17T13:49:39Z

This PR introduces a big upgrade to the CACAO IPLD data structure to make it more concise and compatible with ReCap, UCAN and possibly other object-capability formats in the future.

This work builds upon ucan-ipld by @Gozala and introduces the use of varsig and multidid.

Note that both ReCap and UCAN are currently undergoing some changes in their data structures. This work (I think) represents the latest consensus.

📄 Preview

expede · 2022-12-19T21:32:50Z

CAIPs/caip-74.md

+* `iss` - convert the `iss` string of the UCAN to a multidid
+* `aud` - convert the `aud` string of the UCAN to a multidid


When multidid lands, we'll update ucan-ipld to use it, too 💯

Do we need ucan-ipld if it maps 1-to-1 with the CACAO ipld schema? My intention here was to dedpelicate this work.

expede · 2022-12-19T21:41:56Z

CAIPs/caip-74.md

+type Ability String // e.g. crud/create
+
+type NB { String : Any }
+type Abilities { Ability : [NB] }


I think we need to specify the semantics of this array: it can be interpreted as a logical OR (i.e. "any"), or AND (i.e. "all"). Here's a contrived example:

{ "https://example.com/blog/posts": { "crud/create": [ {"day": "friday"}, {"status": "draft"} ] } }

One can interpret this as either

"You can post drafts to the blog on Fridays"

"You can post drafts to the blog, or post any status on Fridays"

In UCAN <=0.9, these look something like...

// Option 1 [ { "with": "https://example.com/blog/posts", "can": "crud/create", "nb": {"day": "friday", "status": "draft"} } ] // Option 2 [ { "with": "https://example.com/blog/posts", "can": "crud/create", "nb": {"day": "friday"} }, { "with": "https://example.com/blog/posts", "can": "crud/create", "nb": {"status": "draft"} } ]

...which has a very clear intention that the outer array is an OR (or, well, it's using constructive logic, but same thing in this case). Using this logic, to make it an AND, you need to put all keys in a single entry:

{ "https://example.com/blog/posts": { "crud/create": [ { "day": "friday", "status": "draft" } ] } }

I'm fairly certain that without specifying this, we open ourselves to confused deputies based on misinterpreting the semantics from the source capability format.

Good point. Will address this!

Added an example. Let me know if this fully addresses your concern!

chunningham

this ties things together quite nicely

chunningham · 2023-01-05T12:24:52Z

CAIPs/caip-74.md


 ```
 uOqJlcm9vdHOB2CpYJQABcRIgEbxa4r0lKwE4Oj8ZUbYCpULmPfgw2g_r12IcKX1CxNlndmVyc2lvbgHdBAFxEiARvFrivSUrATg6PxlRtgKlQuY9-DDaD-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-PH3T_MZkCeuYK_86RGcbYXRmZWlwMTkx
 ```

+### <a name="appendix-a"></a>Appendix A: Timestamp converstion algorithm
+
+The values in SIWx are encoded as [RFC3339](https://datatracker.ietf.org/doc/html/rfc3339#section-5.6) strings, while CACAO requires unix timestamps (in seconds). The algorithm used to convert between the two is outlined below.


I assume these timestamps can be fractional? or should they be micro/nanoseconds?

They can't, I think we would need to put ms information in the timezone fct properties.

chunningham · 2023-01-05T13:05:50Z

CAIPs/caip-74.md

 Resources:
- {.p.resources[0]}
- {.p.resources[1]}
+- {recap-uri}


perhaps I missed it, but should we explicitly constrain this to one recap-uri per siwx? additionally, if we expect .fct.resources to be readable-ish, should we move the recap uri to the end? users may see a large b64 recap blob and tune out, ignoring the resources after

I think constraining ReCap spec to only include one recap-uri per siwx would make a lot of sense yes!
I don't feel strongly about ordering, but your logic makes sense 👍

I think chunny is onto something with the footgun-factor of big blobs being anywhere except at the end :D

Moved the recap-uri to last.

bumblefudge · 2023-01-12T16:12:23Z

CAIPs/caip-74.md

+* `iat` - is based on `issued-at`
+* `nbf` - is based on `not-before`
+* `exp` - is based on `expiration-time`
+* `fct.z-iat` - timezone info from `issued-at`


Timezones mandatory? Default to 0/UTC if not present for backwards compat simplicity?

I think if there is no timezone in the SIWx and no ms then they can be left out. This will be specified in Appendix A.

bumblefudge · 2023-01-12T16:26:36Z

CAIPs/caip-74.md


 ## Backwards Compatibility

 In the [previous version of this specification](https://github.com/ChainAgnostic/CAIPs/blob/91aaaff73038c2629ff11b88c2209f61521d8ece/CAIPs/caip-74.md), the header type was restricted to `eip4361` as it was designed to work only with Sign-in with Ethereum. Newer implementations should support both header types - `eip4361` and `caip122`.

 ## Example

+**TODO - update these examples**
+
 Below you could find a CACAO, along with its serialized presentation in CAR file.



should "version" be "2" ? Separate CAIP?

"version" is not a property of CACAO, but rather SIWe(x). Realized that there is a "version" section below which I will merge with the backwards compatibility section.

bumblefudge · 2023-01-17T17:04:01Z

CAIPs/caip-74.md

+
+According to the [solana namespace](https://namespaces.chainagnostic.org/solana/caip122),
+
+* `content_multicodec` - set to *caip122*, `0xd510`


but is it caip122 or is it caip122-solana:ed25519 ?

in Haardik's solana/caip122, it proposes the string solana:ed25519 as the signatureMeta, equivalent to eip155/caip122 using caip122-eip191 and caip122-eip1271 😬

solana/caip122 is still in draft so we could change that signatureMeta value... and update CAIP-122 while we're at it before 10 more namespaces get written without a naming convention!

but is it caip122 or is it caip122-solana:ed25519 ?

It doesn't matter in this context I think. ed25519 is defined by the key_multicodec.

It's a good point about cleaning up namespaces after this PR get's merged!

bumblefudge · 2023-01-26T10:08:11Z

Just went to see if the w3c/vc-jwt spec had some language or algorithms that could be cut-and-pasted or referenced for the timestamp mechanics appendices... and found a TODO there as well. well, it was worth a try! I watch that repo so if those sections get filled out before these ones do, I'll drop a link here.

oed · 2023-01-31T20:22:38Z

Moved to CAIP-196 and marked as replacing: 74.

obstropolos · 2023-05-09T19:45:17Z

Bumping on any additional action items here.

Additionally noting recent merge of ucan v0.10 ucan-wg/spec#132

oed · 2023-05-19T09:47:30Z

@bumblefudge I changed this to "Draft", do you think we can merge it so that it's more easily refereneable on chainagnostic.org by implementors?
It would also allow other people than me to follow up with PRs to address the outstanding issues. Both @chunningham and @zachferland are working on implementations right now.

obstropolos · 2023-05-22T14:08:36Z

Seconding - merging the draft is likely a good idea, given implementations and the majority being covered here. cc @bumblefudge

obstropolos · 2023-06-01T15:11:11Z

Since no objection, merging as a draft for now.

Modify CACAO for interoperability

e10ff81

oed requested a review from ukstv December 17, 2022 13:50

zachferland mentioned this pull request Dec 19, 2022

DRAFT: DagJWS CACAO #162

Closed

expede reviewed Dec 19, 2022

View reviewed changes

chunningham approved these changes Jan 5, 2023

View reviewed changes

bumblefudge reviewed Jan 12, 2023

View reviewed changes

oed added 2 commits January 16, 2023 14:56

Address comments

37639e8

Move recap-uri to end

c528936

bumblefudge reviewed Jan 17, 2023

View reviewed changes

bumblefudge mentioned this pull request Jan 24, 2023

AuthZ Biweekly Meeting - 24Jan 2023 ChainAgnostic/CASA#66

Closed

chore: move to different CAIP number

d1cd360

oed force-pushed the fix/update-cacao branch from 1b595c4 to d1cd360 Compare January 31, 2023 20:21

This was referenced Feb 1, 2023

Editorial Meeting 12 Jan ChainAgnostic/CASA#67

Closed

Editorial Meeting 12 Feb ChainAgnostic/CASA#68

Closed

bumblefudge mentioned this pull request Feb 21, 2023

AuthZ Biweekly 21 Feb ChainAgnostic/CASA#74

Closed

bumblefudge mentioned this pull request Mar 10, 2023

Editorial WG Mtg - 9 March ChainAgnostic/CASA#78

Closed

1 task

oed changed the title ~~Modify CACAO for interoperability~~ CACAOv3 Mar 13, 2023

expede mentioned this pull request Apr 3, 2023

First Version ChainAgnostic/varsig#6

Merged

1 task

pedrouid mentioned this pull request Apr 8, 2023

CAIP-222: Wallet Authenticate #222

Merged

Change to draft

db31b92

obstropolos approved these changes May 22, 2023

View reviewed changes

obstropolos merged commit d76ba55 into ChainAgnostic:master Jun 1, 2023

obstropolos mentioned this pull request Jun 1, 2023

Editorial WG Mtg - June 1 ChainAgnostic/CASA#85

Open

mozz30-tech approved these changes May 24, 2024

View reviewed changes

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

CACAOv3 #196

CACAOv3 #196

oed commented Dec 17, 2022 •

edited

Loading

expede Dec 19, 2022

oed Dec 20, 2022

expede Dec 19, 2022 •

edited

Loading

oed Jan 4, 2023

oed Jan 17, 2023

chunningham left a comment

chunningham Jan 5, 2023

oed Jan 10, 2023

chunningham Jan 5, 2023

oed Jan 10, 2023

bumblefudge Jan 12, 2023

oed Jan 16, 2023

bumblefudge Jan 12, 2023

oed Jan 16, 2023

bumblefudge Jan 12, 2023

oed Jan 16, 2023

bumblefudge Jan 17, 2023 •

edited

Loading

oed Jan 17, 2023

bumblefudge commented Jan 26, 2023

oed commented Jan 31, 2023

obstropolos commented May 9, 2023

oed commented May 19, 2023 •

edited

Loading

obstropolos commented May 22, 2023

obstropolos commented Jun 1, 2023

		* `iss` - convert the `iss` string of the UCAN to a multidid
		* `aud` - convert the `aud` string of the UCAN to a multidid


		According to the [solana namespace](https://namespaces.chainagnostic.org/solana/caip122),

		* `content_multicodec` - set to caip122, `0xd510`

CACAOv3 #196

CACAOv3 #196

Conversation

oed commented Dec 17, 2022 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

expede Dec 19, 2022 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

chunningham left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

bumblefudge Jan 17, 2023 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

bumblefudge commented Jan 26, 2023

oed commented Jan 31, 2023

obstropolos commented May 9, 2023

oed commented May 19, 2023 • edited Loading

obstropolos commented May 22, 2023

obstropolos commented Jun 1, 2023

oed commented Dec 17, 2022 •

edited

Loading

expede Dec 19, 2022 •

edited

Loading

bumblefudge Jan 17, 2023 •

edited

Loading

oed commented May 19, 2023 •

edited

Loading