Fwiw, hosting anything on company computers is usually a bit of a no-no. They often don’t like you running anything they haven’t specifically allowed.
I remember in the dot com days, one of the support folks was running a httpd server on his machine. It was rumoured he was giving his mates free websites or some such (back when it was expensive and new!). He got dismissed.
It would be great if nodes can run as many places as possible, but much like mining bitcoins on company machines wasn’t tolerated, I expect the same for autonomi nodes.
Ofc, if you don’t work in a security or IT focussed business, you could probably get away with it… YMMV!
It really may not be a “problem”. They have less connections and see less of the network but certainly still earn, they earn less but I think the amount of cpu saved trumps the loss in earnings.
In other words, it looks ideal for devices with lower resources. Cup half full and all that, certainly not the intention I know.
How about if the nodes seeing the 35K nodes in a network that is at least 75K nodes large would need a quorum of 5 nodes for something, but are not able to contact those 5 nodes?
But is that actually a thing anymore now that we don’t use the native currency?
But would this limited view still lead to unnecessary replication of chunks? Because they don’t see all the nodes, they may think that there’s only, say, 3 replicas and more needs to be made?
Interesting thoughts, idk but to add to that 35k is the average off what all the nodes on a given machine see. For instance at a quick glance one node shows 11k network but it is earning.
p.s. any info on when safenet is gonna be ipv6 capable at last? too bad ipv6 stack was ditched back then when david did his little questionnaire here on the forum
I still have that up my sleeve. Most mobile phones are all on ipv6. Starlink etc. are to. There are ip4->6 tunnels and IPv6 means the end of NAT and we can all work from home with direct client ↔ Client connectivity for an actual fully decentralised messaging system.
As the team are ploughing and steaming to launch I am on the sidelines looking at a few interesting angles we can take here. I hope to publish some ideas in due course, but a mobile app that is ipv6 with even our nodes on ipv4 where we can publish contact ids that clients can directly connect and so on. Then slowly get more nodes ip6 compatible as well.
It’s a process but I think there is a route. First a network though for sure.
Master has some new API changes just merged this week. I’ve not had chance to try them out yet, but I’m assuming they’re the basis for new Python bindings too.
You are not alone with these thoughts. Many here are unsatisfied with the change in ethos of the project. Ask @happybeing
Nowadays it’s okay to introduce paying for safe space via a shitcoin token aka ETH. Something which would have been strictly rejected back in the days. They could have at least tried to build on a modern technology like Bitcoin taproot assets.
Anyways, yes, it’s a hopefully intermediary step. Native token is still planned, although it’s questionable when this will be achieved.
Dang so its “hopefully” an intermediary step? What else has been compromised on, then? Is the network still decentralized? Nodes? Elders? It was supposed to be a blockchainless currency as one of the core main drivers of demand
The native token is being talked about frequently though, as it will unlock much cheaper, smaller and faster uploads. In turn, that brings many more use cases to the network.
Take a look at some of the new data types being discussed and exposed through the API. The native token will likely use a combination of these, and maybe some others.
Edit: To clarify, there are other internal design changes, but not compromises like the token. Just better ways to achieve the goal.