Varun Srinivasan on Warpcast

Content pfp

0 reply

21 recasts

21 reactions

Varun Srinivasan pfp

Varun Srinivasan

The Frame Transactions spec is ready for feedback! Please share your thoughts in replies, comments or DMs. Your feedback is very important to help us get the design right. https://www.notion.so/warpcast/Frames-Transactions-Public-Draft-c2e0d3210d684b4cb7803de1810db36d?pm=c

52 replies

121 recasts

562 reactions

Varun Srinivasan pfp

Varun Srinivasan

Our goal is to collect feedback over the next 24-48 hours and then finalize a spec before the developer call on Thursday. h/t @horsefacts.eth for the heavy lifting and @deodad @gt and @sanjay for design input.

0 reply

2 recasts

37 reactions

Cameron Armstrong pfp

Cameron Armstrong

actually quick question - before going live can y'all make a hidden/dev page in the warpcast mobile app that acts as a playground for testing transactions/how things appear/etc Feels more important now that the stakes are higher than just raw frames

2 replies

0 recast

26 reactions

six pfp

So on mobile, users need to have the Coinbase Wallet app for it to work?

2 replies

0 recast

4 reactions

Shun Kakinoki pfp

Do you guys have plans to support for ‘eth_sign’ for wallets doing account abstraction?

1 reply

1 recast

7 reactions

Nico.cast🎩 pfp

Thanks @v for another comprehensive doc! Can Warpcast/Farcaster protocol maintains a list of “trusted” e.g whitelisted domains publicly, in GitHub? This is more decentralized and 1. helps all devs implementing frames decide which sites to allow 2. allow pull requests for new apps to be added to the list

1 reply

0 recast

8 reactions

ivan pfp

using fctransaction:// scheme would make it possible for wallets to support it with no external dependencies universal links can be specified in farcaster client to use a specific wallet when several wallet apps are installed using fctransaction:// scheme will make it possible to sign on desktop with no walletconnect

1 reply

0 recast

5 reactions

Henri Stern Ꙫ pfp

Henri Stern Ꙫ

Your excalidraw game is on another level @v !

1 reply

0 recast

2 reactions

Stephan pfp

my hot take is that transaction urls should not be in the spec. adds a lot of complexity to what should be relatively straightforward tool for frames to use and is a centralizing force only domain attribution and who shared the frame should be presented to the user in client - no tx simulation, wallets already to this

1 reply

0 recast

2 reactions

jacopo pfp

since trusted domains can unknowingly allow malicious transactions, doesn’t it defeat the purpose of maintaining a list?

1 reply

0 recast

2 reactions

jon wong pfp

in case it is useful, this concept has been battle-tested on Solana—check out [Solana Pay transaction requests](https://docs.solanapay.com/core/transaction-request/overview) in case that provides any useful insight. similar strategy to encode the URL, although given Solana's account model has different params

2 replies

1 recast

7 reactions

Henri Stern Ꙫ pfp

Henri Stern Ꙫ

Left some notes around transaction format and CAIP 10 as well as how to set up UI standards that can work across FC clients so detecting malicious frames becomes easier for users. In general would be very useful to have a Frames reporting site to quickly flag malicious ones

0 reply

0 recast

7 reactions

7858 pfp

POV: You're having drinks with the homies at /ethdenver and someone mentions that the Frame Transaction spec just dropped

0 reply

0 recast

6 reactions

JB Rubinovitz ⌐◨-◨ pfp

JB Rubinovitz ⌐◨-◨

We should have an redteam contest for this on Farcaster

0 reply

0 recast

4 reactions

Michael Silberling pfp

Michael Silberling

Made some comments in the attribution section. cc /data nerds @ilemi @juicemanjames.eth @chuxin

0 reply

0 recast

4 reactions

Volqa Inc pfp

Read on mobile, had trouble commenting. Great work, some feedback: - Consider adding optional authorized_hosts to transaction JSON; limit Frame URLs that can request - Could you support follow-up handlers to "reverts" from transaction, something akin to ERC-3668

1 reply

0 recast

1 reaction

Mark Carey 🎩 pfp

Mark Carey 🎩

Will the call from client to transaction intent URL be a GET or POST? Seems like a post makes most sense, including a similar payload to "post" actions.

0 reply

0 recast

1 reaction

Mark Carey 🎩 pfp

Mark Carey 🎩

In terms of the Warpcast allowlist, something to consider is "posting a bond" or paying in Warps for consideration, much in the same way that you have to pay 100 Warps to post in the /frames channel -- this may reduce the human vetting required by reducing applicants to serious ones.

0 reply

0 recast

1 reaction

Salvino Armati ↑ pfp

Salvino Armati ↑

built different

0 reply

1 recast

21 reactions

David Furlong pfp

Not a fan of this proposal. It adds a lot complexity to the spec and doesn't meaningfully improve over the current problems of crypto UX: speed, safety, simplicity. The flow has significant friction and low conversion for the most common txs in social: small mints/fees, so those txs will find other ways to convert.

1 reply

0 recast

3 reactions