how satellites can be decentralized with only one owner? Fake-decentralized satellites.
There is a difference between technological decentralization of components and decentralization of ownership. Yes, currently the first applies and the second does not. But technological decentralization is important because in the future ownership will be decentralized as well. Does that make sense?
Btw, from what I can tell the stefan-benten satellite is already managed differently from other satellites. So in part this is already happening, but still inside Storj labs.
No just in general not pointing fingers at you or Odmin but with the recently things going on with the payout for some people it seems to bring the worst out,…Just alot of negativity then this cheating with satellites makes no sense, what about all the people cheating storj?
I see what you mean. New SNOs are distrustful. And why shouldn’t they be? They’re dealing with a new entity and they think they have been mistreated. It used to be a lot worse in the early V2 days. I think adding more info on the dashboards is going to help a lot and it’s one of the main reasons I built the earnings calculator to begin with. A black box around payouts is always going to lead to questions that can get heated since it’s around compensation.
But I don’t blame people for not trusting Storj labs during the first months. I trust them, because I have years of experience of them paying me for the services I offered, some times several times over with surge payouts. But why should any new person know that? Just like Storj labs vets every new node, every new SNO is vetting Storj labs as well. It’s just part of the process and we more experienced users should aim to help them in that vetting process. I think the more information they have available and the better the tools are to determine reasonable expectations for payout, the more reasonable this debate will become.
That is true, I trust storj but because I started early. But I see people with more questions then answers I see alot of new people wanting to know why there nodes arent making them tons of money and they signed up for storj thinking they were going to be rolling in the dough. Making storj look like the bad guys here by not creating more data flow for what they have expected it to be.
I get it there’s always debate about if storj is doing the right things but what it comes down too is storj isnt out company we don’t control what they do or don’t do were here for the ride. What bothers me is these people who spend all this money thinking there going to be making tons of money and go rent a bunch of VPS from tons of datacenters and then come to the forum asking why there nodes aren’t making them money.
Storj has to do what they need to do with satellites in general I understand they want to add more satellites to the network, and of course our nodes need to gain trust with them. Storj isn’t just gonna add satellites to the network to make it so we never get our 100% pay.
Because Storj is the odd one out of most crypto related things. Most of the time you get coins by mining and the more hardware you have the more coins you will make. If buying a $1000 miner makes me $1100, that is $100 profit and I should buy 10 or 20 of those miners to get $1k-2k profit.
So, that thinking results in getting entire racks for Storj, which does not work like that at all.
At the same time, at least for me, there is a lower limit of earning that I consider “worth it”. For example, let’s say I have a miner that uses $100 of electricity per month and makes $101 of coins. It would not be worth for me to run it, because the $1/month is not really worth the effort.
So, I did buy hardware for Storj, the idea being that if it takes off, I will get more money and if not, well, I guess I can use those drives for my own data. Otherwise it would mean capping my earnings to about $5/month, which would not really be worth the effort.
I understand that feeling very well. But I think this is partially on Storj. If a new SNO wants to find out what they could earn, one of the first links on storj.io is the earnings estimator. In fact it’s pretty much the only thing that gives any indication. And as of now, it’s simply not realistic. Your post convinced me to turn my alternative that I posted several times in other topics into an idea that can be voted on. I think this should be fixed. Because I once again can’t really blame new SNOs for having the wrong expectations if they are based on an official estimator.
You can find it and vote for it here: More realistic earnings estimator
Edit: This discussion kind of got off topic. So lets discuss earnings expectations further there or the general attitude of people on the forum in a separate topic.
I think I should clear some air about something.
Let me give you 2 examples:
Let’s assume I am a senator and government comes to me for a vote on
Fake Bill 2020. I tell them “I don’t like how point 7 is structured and phrased but I agree with rest of the 130 points. So in its current form I cannot vote for it.”
Someone makes a facebook post and says “Let’s not vote for this president in upcoming elections for reason 1, reason 2 … reason X. But wait who is going to know that we didn’t vote so lets blow up some institutes so they notice us.” Anyone who likes this post is an accessory to a crime as per Cyber Laws of most countries even if you agree only with the first part about
not voting for current president.
My point is, voting/liking is a binary choice. You cannot half like or half vote on something. Personally I would have answered OP’s post like this:
I agree with everything you said but if you can rephrase or remove the
cheatingpart of your post, you have my vote.
This isn’t the first time a company is accused of cheating and won’t be the last. But Storj has been transparent from the start with:
- quarterly earnings
- answering Q&As
- actual names of people working for the company and not anonymous names
- open source code with an invite for anyone to join and chip in their ideas and add their code
I definitely can’t agree with the way this post is openend and some of its arguments and although I felt like I would agree with old nodes not having to go through the process of lowering their held amount, I suddenly realized (while writing my idea) that it is simply not possible!
Let me explain:
STORJ spins up a new satellite and the new satellite wants to keep reputation of old nodes.
This creates several problems:
- where should it get that data from? each satellite has its own database
- your node could be doing great on stefan-benten but is DQed on europe-west and us-central for losing data --> Why should a new satellite simply trust your old node that isn’t as reliable as expected?
- your node could have run GE on some satellites --> point1: where to take the data for reputation
- your node could have removed every satellite except one from the whitelist after first contact. Meaning reputation among all satellites would be great and age old because you never received data from them. And the one satellite you didn’t remove might have DQed you for losing data. --> So even if the data was acquired from all satellites and some kind of average was used, this faulty node would receive a good reputation on the new satellite
- If all satellites are queried and the node only receives a good reputation if its reputation was good among all satellites and has recent activity (as to not cheat the system), it would be unfair to those that did GE or removed satellites from the whitelist
- if your node is great, no GE, no DQ, etc… I understand that it is desirable to not wait 10 month again for full payout.
But in all seriousnes, this is an absolutely ridicuolous edge-case and only a problem because over 90% of the traffic is still test traffic coming from a single satellite. (And the held amount isn’t lost, you still get it and can run GE after 6 month if you wish…)
If we had more customers on different satellites, it wouldn’t matter at all if you have to wait 10 month on a new satellite because you would already have a lot of data from older satellites and still get a lot of data from these. And if these 10 month bother you, you could simply remove that satellite from the whitelist and spin up a new node just for that satellite…
So let’s stop thinking short-term about making the most money in the testing phase and think about ideas that actually make a difference in a real production environment.
hey @anon68609175 Youre bringing up some really important thoughts here. But it seems that some peole on the threaad are having strong reactions to your use of the word “cheating” .
And i think these reactions to the word are adversely affecting the quality of the conversation we could otherwise be having.
Is there a way to express your thoughts which is less distracting to the overall conversation?
I think this conversation is valuable. Thanks for starting it.
I agree, and why even though i’d like to argue for a startup-shared reputation based off of existing nodes on other Sats. It just doesn’t seem realistic to put this type of effort into something like this at the current state of the project.
It would communicate with it’s other peer sats. If ANY other sat has the node DQ’d, or GE’d it’ll start from scratch on the new sat. If it doesn’t match unhealthy criteria it would take the average, or the longest age.
Thinking about this further now though, one could exploit new Sats to their own advantage and GE on them, or fail, and not have any collateral to pay with. So, i don’t see how it can happen.
Fixed. Sometimes I get very emotional when I try to describe ideas
lol understood <3
I think these problems easily be solved like this:
- A new satellite decides what other satellites it trusts. For Storj-operated satellites, this could just be all other Storj-operated satellites. For community-operated satellites, this could be some/all Storj-operated satellites, some other community-operated satellite(s), or no satellites at all (the current situation).
- When there is conflicting reputation information for a given node (there are significant outliers) then the satellite could do one of several things:
- Take the average reputation.
- Take the minimum reputation.
- Not take any reputation and start with the node unvetted (the current situation).
- This “reputation migration” window would only be open for the first one or two months that the satellite begins operation. Beyond that, reputation migration could be used to game the system.
from the eyes of someone very new to all this, seems very unfair this new satellite thing, when the owner of these satellites is the same. Furthermore, can´t see what has to do with decentralization, one more time, same owner. Seem counterintuitive. Reputation sounds like an intrinsic quality of the node. If it´s been catching up with all the 24/7 rules for so long.
Of course, there could be tons of things I´m missing, just wanted to give some perspective on how it looks to someone who is looking to a lot of months of compromise to maintain an infrastructure so demanding, and so risky (just a few hours out and you are again in the starting line after 1 year of efforts).
thanks a lot for the discussion, very enlighten
It’s not just about reputation. Satellites managed by someone else still need to hold back payout in order to be able to pay for repairs should the node suddenly disappear. In fact from what I can tell the only reputation that currently matters actually starts with perfect marks.
Ok so, satellites owned by the same entity. Lets say we make them share held amount in some way. What do you want to do when a satellite is decomissioned? Can’t pay back the held amount, since other satellites still need to be able to cover repairs if your node disappears.
I don’t think the additional complexity is worth it.
This topic was automatically closed 3 days after the last reply. New replies are no longer allowed.