Content
@
0 reply
0 recast
2 reactions
YB
@yb
AMA with @horsefacts.eth starts now! What questions do you have for him?
26 replies
11 recasts
18 reactions
Daniel - Bountycaster
@pirosb3
@horsefacts.eth thank you for doing an AMA 1/2 Farcaster's current design requires users to pay for storage to preserve their casts. Running out of storage not only impacts the OP but also affects those who find these casts valuable.
1 reply
0 recast
0 reaction
Daniel - Bountycaster
@pirosb3
2/2 What are your thoughts on this design aspect? Do you plan to reconsider or modify the storage system in the future? Thanks
1 reply
0 recast
0 reaction
horsefacts
@horsefacts.eth
I wrote a little bit on this back when we published the FIP. Storage is the most experimental of the contracts and we plan to take what we learn from this iteration and evolve it. But someone has to bear the costs, and it's hard to provide public goods. https://warpcast.com/horsefacts.eth/0x0bc23ff3
1 reply
0 recast
1 reaction
horsefacts
@horsefacts.eth
Don't want to seem dismissive of the problem here, though: it is the case that expiring storage has network externalities that impact others and if we can design for it, we should! Do you have ideas here?
1 reply
0 recast
0 reaction
Daniel - Bountycaster
@pirosb3
Thank you so much for the answer and you are not being dismissive at all! The design space is hard and we are all trying to navigate it, I appreciate the FC team for pushing it forward. The most realistic situation imo is to rely on third parties (contractually) to be responsible for long term offline storage
1 reply
0 recast
0 reaction
Daniel - Bountycaster
@pirosb3
While Hubs that are responsible for liveness of the apps commit to keeping unexpired storage, we can rely on third parties to store that data in BigTable so it's still available (but not suited for realtime apps)
1 reply
1 recast
0 reaction
Daniel - Bountycaster
@pirosb3
Putting out a grant for 2-3 teams that commit to independently warehouse all casts and make it accessible would be useful
0 reply
0 recast
0 reaction
