You are viewing a single comment's thread:
Nice work on the site! It looks good, even if it's a bit too AI-overpowered for my taste ;)
I do want to push back on the beneficiary framing in the guides, because I think it misses what that number buys and it sits awkwardly next to "not to pull anyone off 3Speak" in this post. An article subtitled "why 3% beats 11%", sitting alongside "Best 3Speak Alternatives" and "3Speak Down or Not Working?", reads like exactly that.
On the numbers: the 11% is 10% to the platform fund plus 1% to the encoder, and that 1% only rides along on uploads that actually get transcoded. For 3Speak Pro subscribers both are dropped, so it's 0%. Worth saying, since your guide notes igsh can be 0% for some creators too. And that fund isn't purely storage. It covers running the whole thing. But storage and bandwidth are the part that never goes away. Rewards are a one-time event. Transcoding is a real load on encoder nodes, and then the file has to stay there: replicated, backed up, and fast enough to actually stream and seek. That's a bill that arrives every month, forever, and grows with every upload and every view.
Which brings me to the line in your IPFS guide: that these platforms can pay creators more because "hosting costs are tiny", since "the heavy lifting is done by open networks". I don't think that's what's happening. A CID is a name, not a copy. Content addressing says how a file is identified, not that anyone keeps it. Unpinned blocks get garbage-collected and the CID resolves to nothing. Plain IPFS has no incentive layer, so nobody replicates a stranger's 2 GB video for free. Popular files get mirrored; a back catalogue, which is most of any archive, gets zero requests and zero volunteers.
And your own docs answer who covers it: creators bring their own free Pinata key, and igsh.tv "has no server of its own". So hosting costs really are tiny for igsh — because they aren't igsh's costs, they're the creator's. That's a legitimate design choice, but it means 3% and 11% aren't the same kind of number. 3% buys the front-end while the creator supplies the storage; 11% includes the transcoding and the hosting. Comparing them without saying that makes the gap look bigger than it is.
There's a second thing that follows from it: Pinata is one company, and it has nothing to do with Hive. If it changes its free tier, gets acquired, has a bad enough outage, or simply shuts down, every video that only lives there goes with it. The CIDs stay perfectly valid and resolve to nothing. That's a single point of failure sitting underneath a platform whose whole pitch is that no gatekeeper can take your content away, and since igsh has no server of its own, there's no second copy to fall back on. Free tiers are also the first thing a company changes when it needs revenue, and that decision would land on every creator at once.
Your beneficiary article actually puts it well: "running a video front-end costs money: IPFS pinning, transcoding, bandwidth, and development". That's the sentence I'd reconcile with "hosting costs are tiny".
So the honest question is durability: what happens when a creator's free tier fills up, they stop renewing, they walk away, and the key goes stale? Or what if Pinata changes the deal for everyone at once? On 3Speak, that's our problem to solve, on our own infrastructure. On igsh it looks like it's the creator's, on someone else's. I'd genuinely like to hear how you handle that, and whether a video uploaded through igsh will still play in a year. And maybe frame your future posts more honest.