It’s not supposed to be a hybrid. It’s supposed to be one or the other. You either run the services system-wide or in user-mode. For system-wide, you create the service using sudo, but the process launched by the service runs as a safe user, which is non-root. As the post above mentions, the user-mode part was introduced for use with the launchpad, because we didn’t want to prompt people for passwords.
Right now, the node manager makes the distinction between system/user based on whether it runs with sudo or not. The confusion results when users initially use sudo, then with subsequent commands, they sometimes forget. This mistake is inevitable. Prior to the introduction of user-mode services, you would have gotten an error, telling you to use sudo. To support user-mode services, that was removed. From a UX point of view, we had some discussion in the linked thread as to how to deal with it. Haven’t had a chance to work on it yet, so still open to suggestions there. My favourite suggestion was having some kind of command you run which changes your mode from service/user and all subsequent commands will operate in that mode, until it is otherwise changed.
interesting no errors, so somebody earned nanos for that file, than chunk get replicated to my node, than his node gone bad or sth and I am doing all the hard work now, yes, with no earnings at all.
I am not here to earn on safe network I am here cause of taking down centralized systems, that’s the aim, just trying to point out question between lines. The question is: If for example somebody gets rewarded for storing data with nanos, and than his node gone bad or he is going to i don’t know shut it down for some reason, could he before shutting down transfer his nanos to different address and then just shut down his nodes and keep his earnings that way?
I just want to be clear, that way someone could be here for beta wave 1, 2, 3, gain and then get lost? And folks with little or no earnings will do the hard job?
People are getting rewarded for doing the hard work if setting up early, and for helping test and improve, and get to the point where others can join and earn without having put in that hard work.
There’s no unfairness here. It’s just what interpretation you choose to make.
Understood, nobody talking about fairness here, but surely I am the one who asked the question which is interesting to the broad circle of people. Had to take different points of view on that matter (either I, either the creators of autonomi), i believe that others did the same. I am running nodes and even am not registered for the 1st wave so, just speaking hypothetically.
No problem. I think you were asking in a way that implies a view of unfairness so that’s why the responses were directed towards you and why I used the word.
The problems we have are not annoying, because that’s what the network test is for, to expose and fix any problems and bugs, while it’s hard to participate in testing when nothing works and no one addresses any of the problems presented, I asked a lot of questions and no one answered except Chris (but when I wrote to him I didn’t know he was on vacation).
All we need is simple guidance and any answers to questions, and this is lacking.
I’m still looking for the reason for such high resource and energy consumption with only 5 nodes running, and I’m wondering if this could be influenced by the fact that:
The starter by default is installed on the primary memory C:, and the path that I ran I have on another memory with free disk space E:,
when I add new nodes then retrieving latest version for safenode runs on: C:\ProgramData\safenode-manager\downloads\safenode.exe, while the nodes run on: E:\nodes\services\safenode1\safenode.exe,
Could this cause some kind of conflict and as a result a problem with a large load on the computer?
The new node version aims to spread the load of relay connectivity (aka hole punching aka nodes from home), away from the bootstrap peers. This has been causing significant load on those nodes, slowing down all sorts of things. Now nodes will never attempt to relay via bootstrap nodes, hopefully pushing this functionality+load into the wider network.
That’s great! I’d like to connect my nodes via the --node-ports option rather than --home-network so rather than adding to the burden of the relays I’m hopefully taking some of it. But is there any idea of how much extra load that would place on a network connection?
Kind of hard to gauge right now. Since the bootstrap nodes have been taking 100% they’ve been having a tough time of it. So it really depends on how the spread of relaying pans out.
