Content pfp
Content
@
0 reply
0 recast
0 reaction
David Furlong pfp
David Furlong
@df
Considering different decentralized storage options. IPFS vs Arweave vs L2 Onchain Which would you use today if you had to store small amounts of text metadata for a project?
18 replies
5 recasts
32 reactions
David Furlong pfp
David Furlong
@df
If arweave, are you using irys (prev bundlr)? If IPFS are you using Filecoin, ceramic or something else?
1 reply
0 recast
2 reactions
Nero One pfp
Nero One
@n1
how small? if below 1kb i’m going onchain, other than that i’m on arweave.
1 reply
0 recast
2 reactions
Mikko Ohtamaa pfp
Mikko Ohtamaa
@moo
IPFS is not a persistent storage, but a cache/CDN. Think it more of BitTorrent.
1 reply
0 recast
0 reaction
Justin Hunter pfp
Justin Hunter
@polluterofminds
Super biased answer, but IPFS. I struggle to trust the promise of forever when you aren’t the one ensuring that longevity. With IPFS, you control how long the content is available.
1 reply
0 recast
6 reactions
Brenner pfp
Brenner
@brenner.eth
What’s the upload pattern? Whats the query pattern? Do you *really* need to decentralize it at this point? Are your users requesting it? I used IPFS for stuff for awhile and switched back to S3 + cloudfront
0 reply
0 recast
4 reactions
dylan pfp
dylan
@dylsteck.eth
IPFS is appealing as long as there’s a way to query your content — eg I know you can query across Arweave nodes for identifiers within content, but with IPFS you need to know(and probably store) the CIDs of your files
2 replies
1 recast
2 reactions
MetaEnd.degen — ᖽ🎩/dau pfp
MetaEnd.degen — ᖽ🎩/dau
@metaend.eth
arweave definitely
0 reply
0 recast
3 reactions
‎ pfp
‎
@king
It would be Arweave, set and forget.
0 reply
0 recast
2 reactions
Rafi 🍋 pfp
Rafi 🍋
@rafi
IPFS if data is dynamic and structural
0 reply
0 recast
1 reaction
عمر.eth pfp
عمر.eth
@omarperacha
One approach that can enable on-chain storage is to represent the metadata as semantic triples, if it lends itself to that. Idea adapted from EIP-6239 You store much of the text as refs to indices of some mappings, not as text. Then you have a read function that recomputes the full text from the various mapping idxs.
0 reply
0 recast
1 reaction
mmurthy pfp
mmurthy
@mmurthy
We went through this exercise few months ago for our product. We decided to just put it all on L2 to make everything simple. We have large blobs of text now and so are considering Arweave. If really small amounts of text metadata and you are ok with just few more cents/txn, I say L2! UX will be simpler too.
0 reply
0 recast
1 reaction
Greg pfp
Greg
@greg
IPFS for same reason as Justin
0 reply
0 recast
1 reaction
Ben 🟪 pfp
Ben 🟪
@benersing
Why onchain at all?
0 reply
0 recast
1 reaction
CRYPTO GIRLS 👒 🦈 ↑ pfp
CRYPTO GIRLS 👒 🦈 ↑
@cryptogirls
I see that many Web 3 Social projects like Lens or Mirror etc use Arweave
1 reply
0 recast
0 reaction
Jon "JonnyRingo" Williams⚰️ pfp
Jon "JonnyRingo" Williams⚰️
@jonnyringo.eth
Arweave. For small amounts of data it is hard to beat the value it brings. And despite doubting Thomas's, it is built to be permanent, and more reliable than just hosting your own pinning node on IPFS.
0 reply
0 recast
0 reaction
kevin pfp
kevin
@kevinoconnell
Arweave because once i upload it i never have to worry about it again
0 reply
0 recast
0 reaction
sean pfp
sean
@swabbie.eth
Are you considering onchain because potential onchain actions could be based on the text?
0 reply
0 recast
0 reaction
Eric Kuhn pfp
Eric Kuhn
@ekuhn.eth
Who pays to keep Arweave content online ‘forever’ ?
2 replies
0 recast
0 reaction