Ok, that seems like a valid point actually. The generation of safecoins could get choked. It’s crucial that the farming reward continues to function effectively. Would it be possible to reward farmers both for GETs and PUTs?
4 Likes
The immediate problem I see (in an economic sense; there may be technical issues with privacy here that I am not aware of) is that if you reward someone for a PUT, it only happens once. A farmer may get paid for a PUT, but the (spam) data will be there in perpetuity and never accessed by a GET, so this would only make a very small difference to the scenario.
1 Like
The Ponzi argument was raised several times in the other topic (about using SAFE network for free) and his counter-argument was that regardless of the amount of spam uploaded, there will always be some GET-ing going on. That way even if you have only 10,000 GET’s a day, farmers would earn 10,0000 SAFE coins. On a network with a small amount of spam you may have 1,000,000 GET’s a day, and in that case the farmers would still earn 10,000 SAFE coins. No difference!
The problem is that the amount of useful work done by the network would be less. How do we know? Because by not charging for PUTs there would be no condition that data uploaded to the network have any economic use for the network.
2 Likes
I read in some other thread that many GETs only generate a farming reward the first few times a file is accessed. Then caching takes over and that prevents farmers from receiving more safecoins. So effectively there is the same one-time reward issue even for GETs I guess.
I don’t know which topic was that, but caching isn’t permanent (it would be crazy if it was). Caching might take over after 1 or several accesses, but it would also not last forever - if noone accesses the chunk again for a while (say, 24 hours), it will most likely be allowed to expire from the cache and then again it has to be obtained from one of the 4 replica-holding vaults.
2 Likes
But a popular file would likely be accessed again within 24 hours and that would refresh the cache (EDIT: Clarification: refresh the cash timeout instead of fetching the data again), or?
Yes, but it may be popular only for a short period of time (say, photo from a sports match that just happened).
The parameter is easy to change, so it’s up to the devs to decide how they want it to behave. Probably they’ll start with quicker expiry and increase it as they learn more about the way the network behaves.
Whoa, wait a minute, a bit off topic but how can the SAFE network software be upgraded? Wouldn’t that require that all farmers install new software manually? Yikes, that seems like a messy problem.
No, it actually goes quite fast, the network always has 20% of free space. That’s the “base level” of the network, 20% of the space is unused. So with 10.000 users, who all share 120 GB, that’s 240.000 GB of free space. Now a corrupt party starts to upload loads and loads of data, the free space in the network will fall to say 18% so the Farming Reward will go up immediately. Probably even exponentially, so a lot of people will provide extra space, new users start to run Vaults etc. It’s very dynamic.
There’s a very big difference between Bitcoin and SAFEnet which you don’t get in my opinion. Not every user on Bitcoin is a Miner, so indeed, with a small network with miners you could do a 50% attack or something. on SAFEnet all users are Farmers, which you can’t take out. And even if you could ddos 1 ip-address, there will be thousands of them.
In theory yes, practically no. The network is build with extra space available as mentioned above. Farming Reward will change quite fast, so there’s incentive for new users to join quite fast.
2 Likes
I hope not.
I’ve been meaning to raise this scenario for a while, but I already raise too many scenarios, so I didn’t.
But, in this case that you described, one could buy SAFE, take down his 8TB vault and that would take 3% of spare capacity offline. Then after the price raises, he could sell SAFE. And so on.
Maybe the answer is everyone will be doing this all the time so things will cancel out each other. I don’t know if that’s a good answer, but unless the network be very large or implements smart mechanisms for handling farming rate adjustments, this will certainly be gamed.
If can put online 1% of network capacity and terminate it after whatever period the network takes to kill the price, I can probably make a decent amount of money that way. If the network is 4 PB large, a handful of large consumer drives would allow me to play this game.
That is why I think it’s more likely that the network will react very slowly and probably discount (reputation, etc.) capacity changes caused by new vaults.
No one really knows for sure. I also think that 10.000 users is a very small number. I’ll hope to see over 100K quite fast after launch. Quite hard to get 1% of the network at that size. And we should see cancellation when people start to pump or dump the price. A great number wants to buy low at the same time another group wants to sell high.
Those gaming theory guys have elsewhere (not related to MaidSafe) speculated that there may be crowdfunded attacks where participants can pseudo-anonymously conspire to attack together and automatically share the spoils. It’s going to be interesting to watch these games played on cryptocurerncy networks as smart contracts mature.
I think in case of MaidSafe farming rate adjustments smoothed and slow adjustments could help diminish gains from gaming the system as the rate would raise slowly at first (25% empty) and faster later on (10% empty) so farmers could add capacity sooner. Maybe that’s what your screenshot is showing, but I can’t rally figure out the letters!
Not immediately. Remember that the farming reward is controlled by the sacrificial data and, to calculate, we use the total data stored on the network. In fact, the increase or decrease of the total free space does not matter while not affecting the saving of this sacrificial data.
@Anders. Are you from the NSA or something like that. I read somewhere that the NSA try to f…k the open source crypto group be trying to give a good lying mess around. I suspect you to be one of them. Go to my home, arrest me and if I don’t give news after some time everyone will know who you are. Well I hope so.
1 Like
Lol, no I think the core developers and the people at Maidsafe would detect a shill. They have been working on this for almost a decade. I have very little knowledge about the SAFE network.
Perhaps a good reason to read a little, before you suggest a ‘fix’ for the network…
2 Likes
But there are so many technical details! I try to grasp the big picture. For example if app developers have to pay for data storage then they will have to develop apps where they have to earn safecoins. Open source developers will think that’s beneath their dignity. And I don’t blame them. It’s sometimes useful to have money as an incentive, especially for business oriented stuff, but for many applications and services, having to pay for data storage will just be a pain in the neck and a wet old money-incentive blanket over all creativity and freedom.