Maybe I’m being a little too crazy for my own good, but I found myself deeply motivated to build this. It could become both a great addition to what 
At the end of the day, applications would rather work with the more familiar S3 technology. Not only does it solve scalability issues — which, of course, should be priority number one — but it also vastly simplifies file storage, retrieval, and overall management.
Most people probably don’t know this, but
Some could argue — and I’d agree — that IPFS is more crypto-adjacent. For those of us who understand the value of immutability, IPFS feels like the obvious choice. That’s why it’s commonly used for NFTs, for example.
Still, we’d be hard-pressed to argue that immutability is the best solution for every situation.
I’d even venture to say that immutability matters far more at the base layer — at the protocol layer — than it does for everything built on top of it. Blockchain protocols store text, simple data, and references to media, but making every image or video immutable forever doesn’t necessarily make sense every single time.
I recently discovered a clever trick YouTube uses to process videos faster.
When you upload a video, YouTube first processes it at standard quality and immediately publishes it. Then, in the background, it gradually generates the higher-quality versions. That’s why your video is usually shareable within a few minutes, while the 4K version might take another thirty minutes to fully appear.
It’s a brilliant approach — and one I considered imitating.
Then I realized IPFS would never really allow it.
The folder structures themselves are immutable, which means I can’t simply append new quality versions later on. I’d have to delete the structure entirely and recreate it from scratch.
And that’s really what all of this comes down to: compromise.
A better user experience almost inevitably demands a more centralized system. Meanwhile, a more resilient system will almost always come at the cost of speed, flexibility, or convenience.
With all that said, working on Butter Cloud — as I’ve decided to call it — and having this technology in our portfolio would allow us to play with those dials a little more intelligently.
Maybe we centralize things upfront for speed and usability, only to decentralize them later for long-term storage and resilience.
Maybe that’s the judo move hiding in plain sight.
At any rate, the coals are hot, and I’m cooking something.
MenO