MaidSafe Dev Update - September 21, 2017 - Alpha 2


HI. I have Trust Level I but for some reason I’m unable to log in to the network using my original credentials. So, I used the “Claim Invitation” option, selected Alpha 2 and the page that pops up remains with “… waiting” message. I’d like to try it out but can’t log in.


Did you set your IP on the invitation page?


I have always wondered what affect “passive participants” would have on the efficiency of the network. By passive participants, I mean people who start storing data when the price is attractive and then go about their business without paying any attention to the price of safecoin and their own vault’s profitability, putting their vaults on auto-pilot, so to speak. If enough vault-owners did this wouldn’t it cause the efficiency of the protocol and the operation of the entire network to suffer?


Why would the network suffer because people are hosting data relatively cheaply? Seems like a good deal for users, no?


What happens, though, if the price of data goes up and stays that way for a considerable time but many people continue their hosting anyway because they are either unaware or unconcerned about the increase in the cost of hosting data?


Could you explain why this may be a bad thing?


If the price goes up you’ll see everybody rushing to host a vault and it won’t stay high for long.


Actually, I meant, what happens if the price of safecoin goes down and stays down for a considerable amount of time. The efficiency of the network depends on a significant number of vault-owners disconnecting their vaults, right, so the price will go back up? What happens if they don’t disconnect? What happens if there are enough “passive vault-owners” (maybe because the cost of hosting is insignificant to them) who stay connected anyway? Let’s say 60% of the vault-owners fall in this passive category. Will the price-setting algorithm of the network be able to adjust and continue to be efficient, lowering and raising the price depending on what the “active” portion (the 40%) part of the network does?


You have to separate the price (set by humans in market) and safe/get safe/put ratios (set by network). In your case, with high supply of farming relative to need, the safe/get paid will be decreasing (decreasing potentially saleable supply), and the safe/put will be cheap, making it cheaper to use the network (more bang for spending). On the fiat side, there is another parallel feedback loop. For the rational actors, buying safe instead of keeping a marginal vault running will look good (adding buying demand), and for new users, the ‘how do I get my first safe?’ will be answered with simply buy it rather than hosting yourself. I think we all hope the network is as cheap as such service could possibly be provided… Having a large number of passive, altruistic vaults, using ‘spare’ capacity, which is the cheapest there is, is probably really important to keeping the network cheap enough for all the interesting use cases. Interestingly, this doesn’t require that the token itself is cheap, just that at whatever price the token commands, the incentives to buy vs add extra capacity to the network are roughly balanced.


Interesting. Thank you for your thoughtful response. My concerns did not revolve around altruism being the reason for passive vaults so much as a lack of interest by the vault-owner. Maybe that doesn’t matter. “Passive, altruistic vaults” being a positive thing for the network seems counter-intuitive to me but I hope it works out that way. I can see now how the interdependence of real markets for safecoin and the price-setting algorithm of the network might work hand-in-hand to produce the intended affect.


I am still struggling with what your fear is? If the PUT price is cheap, it is because there is plenty of storage available. Who is this a problem for and why?


Yes, exactly! I can only see cheap storage as a good thing. The network will have to compete with the likes of Google and Amazon for storage costs and it really wouldn’t be a good start to be more expensive.

If safe net really takes off, I suspect there would be many years of new farmers joining and adding storage to the network at a very low cost. Those users will be essentially getting something for nothing, while other users benefit from cheap hosting - win, win, as they say!


It just seems possible to thwart the dynamics of the internal price-setting algorithm if there are enough “passive” vaults. In my example above 60% of the vaults are passive and the network cost of storage (internal safecoin price) is low so it follows that the result would be that some vaults would drop offline because of the low price, i.e. low profits of running a vault. But the only vaults dropping offline are the ones among the 40% active-owner group, so it is only a churn of a minority of the network. The 60% passive group who never disconnect will keep the price artificially low no matter how many in the active group drops out.

I can see now, however, that the real human markets for safecoin might be able to offset this “slanted” pricing mechanism of the network and tend to raise the price back up if enough in the active vault-owning group drops out of vault-owning and starts to purchase safecoin on the open market. That will drive the price, at least on the real markets, back up. The question remains, though, will the cost of storage stay the same on the network or will the churn in the minority active group be enough to start raising the price of storage again?

It just seems to me we are depending a lot on there being a symbiotic relationship between the real markets and the algorithmic market of the network. We could end up with most, or all, of the 40% active participants dropping out of vault-running and the only ones left are the passive ones. That would mean a constant low cost of storage, wouldn’t it? The price of safecoin on the network would no longer be dynamic, it would stay low and the only dynamic movement would be on the real markets. Is that a good thing? Wouldn’t that possibility lead to an under-supply of storage eventually, which would lead to an even greater surplus of storage later on as more and more vaults are added? What if the majority of these new vaults are also passive? Storage on the network gets cheaper and cheaper while the price of safecoin on the markets probably keeps going up. Is that desirable? Maybe so, maybe that would be the best of both worlds, but I don’t think it would be the way the network was originally designed to work, maybe indicating problems down the road.

This is more of a thought exercise than a fear.


It would seem pretty desirable to me. I wouldn’t say that the price would be artificially low either - an influx of previously idle resource will cause an increase in supply. Making use of idle resource is surely a net gain for everyone though, right?

It may be that dedicated farming will only be profitable waaay into the future. Exhausting all idle resources may take quite some time and may go hand in hand with a shift to storage being considered a local data cache instead of offline storage. Even when dedicated farming does become profitable, it may be in embedded devices before data centres. The dynamics may well be very different to what we expect with tradìtional data hosting.

However, I am sure the day will arrive when the network needs to incentivise new storage to come online. At that point the storage cost would start creeping up*. Maybe the first impact will be people freeing up some space to profit from this. Maybe after then, people will add external drives and the like. After that, maybe competition will be getting fierce and response time becomes a driving factor, feeding dedicated hosting. I am quite certain it will happen relatively slowly either way - there is a lot of untapped idle storage out there!

Edit: * ofc, storage continues to get cheaper year on year. I doubt the cost would ever actually go up for a sustained period of time. I suspect it would just go down slower.


Well, quite interesting and I hope we’ll get soon to a beta. Keep up the good work.


Beautifully written. You know what they say: information wants to be free :slight_smile:


Efficiency of the network IMPROVES as more people run vaults. The profitability of safecoin for the speculator may go down though. Your statement suggests you are saying the efficiency of the network is the efficiency of it making profit for people.

You need to realise that the network would love more vaults to remain on line, and the price to store data would be cheaper too. Think of how that helps the poorer of the world to store their data safely and securely.

Basically it helps everyone EXCEPT the ones after a profit.

The cost in safecoin to store data only increases in response to a decrease in network efficiency due to not enough vaults to maintain optimum performance (space/bandwidth/etc). This encourages more farmers.

EDIT: A question if I may. How do you see the primary purpose of creating the SAFE project. To make farmers profit? Or to create a network that is secure access for everyone? Don’t tell me if you don’t want. I asked the question because it might help you understand why things are done a certain way.


@maidsafe, which version of safe_vault is being used for alpha2? From the commit history it looks like it would be 0.17.2 but it would be good to know for sure in case anyone wants to do some private network testing.


Yes at the moment 0.17.2 would be what you are after for private network testing.


That’s the way vaults and safecoin should work in theory, as designed. Whether it will work that way is another matter. To tell you the truth we will not even know, or even be able to speculate, on how the dynamics of vaults and safecoin will work even when we get to the beta version because users will not have the same motivation or lack of motivation when it comes to operating vaults. The network will need to be on its own for a matter of months before we can get an indication of how that dynamic is shaping up.