Looking at my ISP accounts, they are 99% garbage, and nobody cares.
I see no point in make a platform that does care - as it wouldn’t even compete with what is already out there.
Lots of fuss about nothing.
I agree with you to some extent.
But there is a difference on SAFE. If one vault could dump their data (mostly garbage) onto another vault. They would free up their space for resell. This actually happens on SAFE when a vault is reset for whatever reason.
Reset = disconnected from SAFE, then reconnect as a brand new vault and start over at rank 0.
Data Life Cycle
If 98% data remains “active” for the first 30 mins… then it makes sense to reset the vault every time it fills up. Because the rate of GET requests will decrease as those “active” chunks get older. Eventually you would have a full vault with GETS from only 2% of your chunks. There may be a few gems (frequently GET data) but most chunks will eventually switch to “inactive.”
Problem
Because I care about SAFE’s sustainability, I would rather not have farmers dumping garbage on each other. This could lead to a systemic chain-reaction. IMO that is a potential danger to the Network. And for that reason, it is important to discuss, consider, and prepare… even if it never happens. This is why recycling is being brought up.
Solution
If SAFE does autonomously archive data to the largest high ranking vaults, it reduces the “temptation” for profit seekers to manually reset their vault. This is a good thing. We just have to make sure those landfill vaults are compensated for looking after archive data. And it does this is by giving them a higher farm rate for “active” data chunks.
Landfill Farmer Incentive
Landfill farmers would keep adding storage to just 1 vault, because the (Network Average) will keep increasing as the Network grows. This may reduce the need to recycle… but it’s too early to tell right now. If this incentive doesn’t work, then recycling should be considered.
4 Likes
Yes, to compare only the cost of storing data, ignores all the value add…that the SAFE Network brings.
To imagine a network of personal computers could even compete with ‘cloud storage’ is pretty amazing.
Doesn’t vault longevity effect reputation, and thus profit?
Reseting the vault ought to start fresh with reputation.
Data is always going to get stale. Once you get your new vault’s reputation up you will also be full with staling files. Should balance things out.
I don’t think anyone is going to get rich running maidsafe vaults. The market should balance it out at someplace around a break even level unless there is a shortage.
1 Like
Yes, it does. But you need to get GET requests for it to happen.
We’re talking about data chunks that (essentially) never receive them.
1 Like
IMO vaults will have a peak earning cycle, like a bell curve. As long as new chunks are added, income potential remains the same. But as soon as their max capacity is reached, there will be no new chunks to make up for the old ones. At that point, Safecoin earnings will drop significantly.
This event will affect smaller vaults very quickly. While bigger vaults can run much longer. The only way for smaller vaults to regain their earning potential is by resetting their vault, or having “inactive” data archived to the larger vaults.
Some people will farm on their mobile devices, and some will dedicate big server farms. If we balance out the incentives with data management, recycling may not be needed. Unfortunately, I cannot predict the future so I have to consider everything.
Side Note: this leads me to believe SSD would be ideal for small vaults (Network Average or lower) while HDD is ideal for large vaults (20% Above Network Average)… interesting.
2 Likes
In order to get data chunks that are likely to get GET requests, you will have to accept a lot of chunks that don’t — Nobody will know the difference… And you don’t get to choose — You will get the files you get - and those may or may not be new. Even with a brand new vault, iI your X/OR neighbor turns off his machine, you are going to get his stale files. Nothing you can do about it. The lottery tickets fall where they may.
1 Like
Those are some good points.
@jreighley is passionate about his viewpoint but for those farmers who don’t pay for bandwidth even small differences will matter.
I doubt I will personally care and most single disk farmers probably won’t care too, but those who invest $1K or more in their farms will be paying attention.
True, but a single change in the way rewards are calculated or value of some parameter may screw one’s plan in no time. Pro farmers will certainly want to know the final, production parameters before they commit to buying h/w. Even then it will be guesswork because mainnet workloads will probably be significantly different .
2 Likes