2025 July 16 Community Call

[expede]

Migrating here from ucan-wg/.github#6

🍿 Video

YouTube

📝 Notes

These highlights were pulled from the July 16 UCAN Community Call. Please edit or comment with anything else that was missed.

• We're so back! Lots of great progress lately:

  • Spec has reached v1.0.0!
  • 3 active implementations:
    • Rust (@expede): https://github.com/ucan-wg/rs-ucan
    • JS (@hugomrdias): https://github.com/hugomrdias/iso-repo/tree/main/packages/iso-ucan (with other packages like did and signatures there too)
    • Go (@MichaelMure): https://github.com/ucan-wg/go-ucan

• New website is coming (@cwaring)

  • Staging(ish) link: https://ucan-spec.pages.dev/
  • Aims to pull content from Github
  • Further discussion in Discord

• Tooling and code examples:

  • From Infura/MetaMask/Consensys:
    • https://github.com/ucan-wg/go-ucan is a (pending varsig update) valid UCAN 1.0 implementation with delegation and invocation
    • https://github.com/ucan-wg/go-varsig is a Varsig 1.0 implementation
    • we are in the process of making public go-did-it, a DID implementation (crypto handling included) that will be bolted into go-ucan
    • we also have a bunch of tooling around go-ucan that we may or may not release as well (client side, external arguments to use UCAN wrong)

• Testing: We need to get a standard set of test vectors out there, but until then, let's share test vectors and I/O as much as possible.

• Shared infrastructure for revocations, etc.

  • Still in early stages, but looking for individuals or orgs who want to maintain a node
  • Goal is for software to be very lightweight (fits in free Cloudflare tier, for example)
  • Looking into League of Entropy as an example of collaborative community infrastructure

🪵 Chat Log

09:11:34	 From Hugo Dias : we can probably demo the new website
09:11:49	 From Benjamin Bollen : Reacted to "we can probably demo..." with ❤️
09:12:23	 From cw : https://ucan-spec.pages.dev/
09:13:43	 From cw : https://ucan-spec.pages.dev/llms-full.txt
09:14:59	 From Michelle Lee : will you be using simple analytics (maybe plausible)?
09:16:58	 From Michael Mure : what could be nice is explain how to use UCAN for existing APIs with the least pain possible
09:17:10	 From Brooke Zelenka (@expede) : Reacted to "what could be nice i..." with 💯
09:17:25	 From Michelle Lee : Reacted to "what could be nice i..." with 👍
09:17:37	 From Michelle Lee : code examples would be great - maybe they’re already in the docs and we can highlight?
09:19:30	 From Michelle Lee : there’s some good content that’s slightly more opinionated/informal - Brooke’s talks at LoFi & IPFS events, Storacha Docs, etc. https://docs.storacha.network/concepts/ucan/ 

Is there a way to highlight these as “social proof” or a more independent introduction/persuasion for UCAN?
09:19:56	 From Michelle Lee : also old web3.storage blog post - https://blog.web3.storage/posts/intro-to-ucan
09:24:46	 From Michael Mure : in non-public-yet code, I have some example code to have:
- an issuer/service
- a client that do request against the service
- a client that get a sub-delegation from the other client and query the service

imho that would be a good intro example for the website
09:25:04	 From Hugo Dias : Reacted to "in non-public-yet co..." with 👍🏻
09:25:06	 From Brooke Zelenka (@expede) : Reacted to "in non-public-yet co..." with 💜
09:25:08	 From Rob Elves : Reacted to "in non-public-yet co..." with 👍🏻
09:25:18	 From Alan Shaw : Sorry I'm late
09:26:04	 From cw : Reacted to "Sorry I'm late" with 👋
09:26:08	 From Brent Shambaugh : Reacted to "Sorry I'm late" with 👋
09:26:26	 From Michelle Lee : RIP repeated online events 🙁
09:26:41	 From Michelle Lee : Reacted to "Sorry I'm late" with 👋
09:27:07	 From Brooke Zelenka (@expede) : Reacted to "Sorry I'm late" with 👋
09:27:25	 From Michael Mure : it's great :)
09:27:30	 From cw : Reacted to "it's great :)" with ❤️
09:27:35	 From Michelle Lee : WE CAN!
09:27:40	 From Benjamin Bollen : Reacted to "it's great :)" with 👍
09:27:48	 From Michael Mure : I'm in the middle of my holiday with crap bandwidth so can't really talk but I wanted to mention that (from Infura/MetaMask/Consensys):
- https://github.com/ucan-wg/go-ucan is a (pending varsig update) valid UCAN 1.0 implementation with delegation and invocation
- https://github.com/ucan-wg/go-varsig is a Varsig 1.0 implementation
- we are in the process of making public go-did-it, a DID implementation (crypto handling included) that will be bolted into go-ucan
- we also have a bunch of tooling around go-ucan that we may or may not release as well (client side, external arguments to use UCAN wrong)
09:28:17	 From Benjamin Bollen : Reacted to "I'm in the middle of..." with 👍
09:28:34	 From Brent Shambaugh : I'm late too, but my brain decided to remind me 20 minutes in. Yay, tech call.
09:29:06	 From Michael Mure : YES
09:29:09	 From Michael Mure : PLEASE
09:29:42	 From Brooke Zelenka (@expede) : Reacted to "I'm in the middle of..." with 👍
09:30:03	 From Benjamin Bollen : Reacted to "I'm in the middle of..." with ❤️
09:30:18	 From Hugo Dias : This is the JS repo https://github.com/hugomrdias/iso-repo/tree/main/packages/iso-ucan  with other packages like did and signatures there too
09:30:32	 From David Luecke : Great, I was just about to ask! 🙂
09:30:37	 From David Luecke : Reacted to "This is the JS repo ..." with 👍
09:30:56	 From Brooke Zelenka (@expede) : Reacted to "This is the JS repo ..." with 👍
09:31:33	 From Alan Shaw : We're still stuck on 0.9 😩
09:31:36	 From Brooke Zelenka (@expede) : For Rust: https://github.com/ucan-wg/rs-ucan, and my branch is v1-2025-edition
09:32:21	 From Brent Shambaugh : I modified the did-jwt library for P-256, sans JWE, and mircea was there to help. wrote a bunch of tests... I think this is what you are doing now, but for  UCANs. thinking out loud
09:32:37	 From Brooke Zelenka (@expede) : Reacted to "I modified the did-j..." with 👍
09:34:04	 From Michael Mure : Reacted to "This is the JS repo ..." with 👍
09:38:09	 From Brent Shambaugh : What sort of hardware is needed? Can it run on an ESP32, a raspberry pi, or does it have to be something more heavy? Is there something out in the wild I could run with? (could be a to do for me) thinking out loud.
09:38:39	 From cw : Reacted to "What sort of hardwar..." with 👀
09:38:59	 From Michelle Lee : https://www.drand.love/loe
09:39:49	 From Hugo Dias : Reacted to "I modified the did-j..." with 👍
09:39:50	 From Mircea Nistor : Reacted to I modified the did-j... with "👍"
09:40:05	 From Mircea Nistor : Reacted to This is the JS repo ... with "👍"
09:40:12	 From Michelle Lee : the goal of “it runs on free cloudflare tier” would probably help
09:42:27	 From Brent Shambaugh : Replying to "What sort of hardwar..."

so I could go smaller, with an SD card if storage is the limitation.
09:43:39	 From Brent Shambaugh : Replying to "What sort of hardwar..."

The SD card will store  the revocation list. Even 1TB is cheap for an SD card.
09:44:07	 From Brent Shambaugh : Replying to "What sort of hardwar..."

ESP32 does not run the linux kernel
09:44:26	 From Brent Shambaugh : Replying to "What sort of hardwar..."

ESP32-C6 is an exception
09:44:34	 From cw : Reacted to "the goal of “it runs..." with 👍
09:46:36	 From cw : Replying to "What sort of hardwar..."

Would be fun to try this!
09:47:06	 From Michelle Lee : Does anyone have a link to the Discord?
09:47:15	 From Michael Mure : thanks everyone :)
09:47:24	 From Benjamin Bollen : Thanks for all the great work
09:47:26	 From Brooke Zelenka (@expede) : https://discord.gg/PmRb956S