💾 Archived View for bbs.geminispace.org › s › permacomputing › 11755 captured on 2024-06-16 at 16:42:23. Gemini links have been rewritten to link to archived content

View Raw

More Information

⬅️ Previous capture (2024-05-26)

➡️ Next capture (2024-07-09)

🚧 View Differences

-=-=-=-=-=-=-

thinking about the idea of fault tolerant networks and protocols. There's a certain amount of overhead to sync-style protocols, and one aspect of this is that updates are tracked. They must be in the protocol, otherwise old versions may propagate.

gemini specifically seems to have a gap here, in terms of versioning and proxies.

Posted in: s/permacomputing

🐉 gyaradong [mod]

2023-11-16 · 7 months ago

3 Comments ↓

🐉 gyaradong [OP/mod] · 2023-11-18 at 00:37:

one idea behind collapse computing is intermittent networks, where both core and edge networks have intermittent failure. This means point to point connections are sort of impossible. you instead need to be able to instruct the fabric to send data "later" through proxies or some similar mechanism. IPFS solves this through being a content addressable storage system, so you can grab data from anywhere that has the content regardless of origin. HTTP has proxies. Gemini sort of doesn't.

🐉 gyaradong [OP/mod] · 2023-11-18 at 00:39:

the other side of that coin is that gemini is fairly simple. Ask for data and get it. A small computer can do that. it probably couldn't do the maths necessary for hosting an ipfs style system.

🐉 gyaradong [OP/mod] · 2023-12-09 at 14:42:

idea from scrawlspace: just save the text. hmmmmm this makes a fair bit of sense. Maybe a dated log of things downloaded rather than an opaque cache.