Advanced
Varun Srinivasan@v
3/12/2024

What's the most annoying thing about building frames today?

frames
In reply to @v
Matthew Fox@matthewfox
3/12/2024

not having a local testing playground if you don't use frog

frames
In reply to @v
Vladyslav Dalechyn@dalechyn.eth
3/12/2024

Being not able to invalidate cache of a frame. Upgrading a frame therefore becomes problematic. A button in the validator would be enough for such!

frames
In reply to @v
Corbin Page ๐ŸŽฉ@corbin.eth
3/12/2024

Frame caching can be tricky to deal with

frames
In reply to @v
3/12/2024

Building the dynamic frames with node-canvas. Couldnโ€™t get Satori to work for some reason

frames
In reply to @v
veekay โŒโ—ง-โ—ง@
3/12/2024

lack of debugging tools. lack of clear error messages leading to unable to find the bug let alone solve it

frames
In reply to @v
Emmaniii@emmaniii
3/12/2024

We need transactions to work on test networks (i.e. sepolia) to make it easier to develop and test frames

frames
In reply to @v
Justin Hunter@polluterofminds
3/12/2024

Images for interactivity is a super hacky experience. Sandboxed HTML would enable almost all the use cases people are hacking around today.

frames
In reply to @v
Victor ๐Ÿ–๐Ÿง€๐ŸŽฉ@victoctero
3/12/2024

Not finding the way to build a no-code raffle frame (maybe there is some workaround and I donโ€™t know it)

frames
In reply to @v
Cameron Armstrong@cameron
3/12/2024

- no always up to date before releases canonical page for every frame spec change - constant moving target for spec changes (#1 would help w this) - no testnet for txn - no mobile playground for testing UX flows when 80%+ warpcast users are mobile - frame dev support is doing their best, but there's too much to deal w

frames
In reply to @v
3/12/2024

1. Not having support for a testnet for txn frames 2. A call back endpoint for a normal frame would be really cool.

frames
In reply to @v
Daniel - Bountycaster@pirosb3
3/12/2024

Not annoying anymore since we build a framework to abstract this, but I think this change could be high ROI: Having button actions as enums instead of indices reduce the requirement for state tracking + reduce problems related to backwards compatibility

frames
In reply to @v
maxi ๐ŸŽฉ๐Ÿ–๐Ÿฟ๐ŸŒˆโ˜˜๏ธ@maximal
3/12/2024

425 $DEGEN

frames
In reply to @v
kugusha โ†‘@kugusha.eth
3/12/2024

not building, but using https://warpcast.com/kugusha.eth/0xbe09be38

frames
In reply to @v
Shashank @0xshash
3/12/2024

frames are extremely ephemeral with no great discovery mechanism today a good way to bookmark/ browse recently used frames/ discover new ones would go a long way

frames
In reply to @v
3/12/2024

Back button handling, not as tricky, just extra work to handle, plus sometimes you're simply out of 4 buttons

frames
In reply to @v
froggie@froggie
3/12/2024

* lack of end-to-end examples using modern tooling * we are so early - so all the tools are new and have many bugs * dependency on external services such as syndicate and neynar for common functionality such as verification and fid to address * lack of standard error handling * tx domain verification seems broken

frames
In reply to @v
Fariba Karimi๐ŸŽฉ๐Ÿ…๐Ÿ”ฎ@fatiba1363
3/12/2024

172 $DEGEN

frames
In reply to @v
timdaub ๐Ÿฅ/๐Ÿซฆ@timdaub.eth
3/13/2024

that there is a lot of competition and you donโ€˜t earn much for the work

frames