Using only SafeCoin, just makes it expensive but doesn’t limit the ability of bots to register rapidly many PublicIDs. Any numpty with a few coins can write a bot that would be not very expensive relative to the disruption it would cause. There’s no good reason to reward greed.
You can use SafeCoin as well but preferring vaults that are valuable to the network will strengthen the network. It’s a simple win to ensure the focus at the start is on providing vaults that are useful, rather than being a land grab.
Well obviously, such a simple thought, can be more widely applied. The point of keeping it simple, is about limiting complexity so the intent is not lost in the communication and application of those ideas by or to stupid.
Leaving yourself or a product open to exploitation, just for the sake of being simple, is not a great idea.
I rather like the idea of one free id, then an incrementally larger fee for each additional one. Will make it harder for the bots … I wouldn’t be opposed to a small charge for the first id either. I don’t mind bots in the end, so long as there is a cost for them to join the network. Bots could do good things for the network as well and so long as the cost isn’t completely free, the bad bots will be at a disadvantage.
If we could differentiate between a deposited and a farmed Safecoin, maybe something like:
an account is free to create, and can pay to store data (exchange Safecoin for PUT), but can only create public IDs that generally are not desirable (e.g must start with at least four alphabetic characters, and must end with four or more digits, so “Tyler9999” for example.). This ensures anyone can create a website, and do all things that require a public ID right away, but stops creation of accounts just to squat desirable IDs. So what is the hurdle for creating desirable IDs…
if a vault farms one Safecoin directly into a wallet, the account controlling that wallet earns the right to create N public IDs without restrictions. (Say N=5 to start, but ideally would change as the farming difficulty increases over time). Every time a Safecoin is farmed into a wallet, this number increases (because that means there is no advantage for those able to keep creating new temporary vaults, over your average user without a botnet, server farm etc = level playing field based only on the resources you provide).
This incentivises early farming greatly, and makes it impossible to gain control of lots of nice IDs without significant cost, and ensures the network (ie everyone) benefits from the resources provided by people motivated to get vanity IDs (whether they are farming them, or paying for them and thereby incentivising people to farm them).
The benefit from incentivising people to farm for IDs: more resources (from people trying to register more IDs) reduces storage costs for everyone.
It complicates the incentives for farming, but not in a detrimental way I think, and eventually the effect will disappear as all the more desirable IDs get taken and the incentives reduce. This will I expect happen over a much longer period than the current DNS where large numbers of donations can be bought by anyone with enough cash. This mechanism requires those acquiring these IDs contribute far more in exchange for them than with the current DNS, with the benefits going directly to the network and all its users.
With this system I think it is fine to allow a free market in IDs because that provides the incentive for people to provide storage in order to farm IDs (and Safecoin - they get the Safecoin too of course).
The system avoids the network needing to charge directly for an ID, which is good because pricing them effectively is very difficult. That’s because it is very hard for the network to know which are valuable and which are not. An auction might solve that, but I don’t see a way for the network to implement that. At least that looks like a much harder problem to me.
The one difficulty I see immediately is deciding the restrictions to place on the “free” IDs in order to avoid these becoming desirable for squatters, but I think that’s an easier problem. We will have to look at how these might be valued by users and try to ensure they are ugly enough, and that there are enough of them, to make squatting too costly to turn a profit! Maybe the network can increase the prefix/suffix lengths over time, or… but now it’s getting complicated again
Maybe the suffix has to be longer is my first thought, and maybe it has to have some none-alphanumeric characters too - the aim being too ensure there are lots too many such IDs to make it practical to hoover up enough to make them scarce. At the extreme they would look like a bitcoin address.
Perhaps also there could be a free lottery… allowing anyone to play some game for the chance of a random generated PublicID. 1 play per hour that a launcher is left running or similar; and then 1:14000000 odds of landing safe://google but 1:10 chance of getting something more desirable than the default irritant.
Taking money out the equation is a nice idea.
I expect there’s some happy medium to be found along those lines above.
Not really, first you have to decide what the network does and does not permit, and that is in the network itself. After that you can build things on top, and that’s the app level. What I’m suggesting has to be underneath that.
Initially I had in mind some time limited fix to get over the initial 6months-year problem. Whatever is most simple to address the problem. I don’t see a big issue of it being at app level.
Can someone of the maidsafe team tell us if their still will be many resets of the network, cause im making new id’s for my projects every time but would like to know when i can stop making new ones.
That’s because it’s not clear yet. If the first implementation of every feature is right in one hit, it will not take as long as when it takes ten tries before Safecoin (for example) works as it should.
You’ll probably read it all over the place when Maidsafe knows when the ‘real’ network goes live
It can’t start until all the testing is finished, and that won’t be known in advance because it will be determined by the results of testing, which are unknown until the tests are carried out.
If publicIDs cannot be deleted and everyone needs a unique publicID then what happens when publicIDs are forgotten?
I once read a short story about how there came a crisis caused by severe copyright. That the statute of limitations had been revoked and now people could own copyright indefinitely. It drove the protagonist to execute an assassination mission to attempt to stop and reverse this because if words and stories are forever copywritten then eventually you run out of combinations of ways to retell them. Likewise in our predicament here if unique publicIDs last forever but can be forgotten and never retrieved then essentially that WORD is lost forever and forever locked out of the SAFE network for anyone ever to use again. Sure you can use symbols and numbers but are we really going to reach a point where we have to start naming companies like “FruityFlakes325534352-235523” Just so we can get a public ID?
What I’m trying to do is take this to it’s logical insane conclusion. When IvP 4 was created no one thought they’d run out of DNS numbers. Lo and behold it happened and IvP 6 had to come along and EXTEND that number. The internet got REALLY big really fast. If anyone can make their own public IDs and websites at virtually no financial cost how fast do you think publicIDs will be taken up?
Beating domain, or PublicID, squatting isn’t for the sake of convenience. It’s a survival necessity.
Good point, and another reason why charging a small rent on public IDs would be desirable. When the rent goes unpaid they would be released after a grace period.
A petname system doesn’t provide equivalent and valued functionality, which is why hardly anyone uses it. It doesn’t deliver universal addresses, which is one of the most valued properties of domain names and why so many people value and use them. If it was as good a solution I’d be all for it.
Here, rent is used in a different way completely, and it seems you haven’t understood the proposal. Try understanding the use of rent here before commenting.
What if we have something that’s more like a cycle/recycling process, sort like in nature (if we continue the farming metaphor).
To me, “rent” is paying x amount every y unit of time.
Instead, what if we pay x amount upfront for a y unit of time. After y passes, the fee you paid is returned to you and the publicID is now up for grabs again.
Pros:
It’s not like rent, which solves @anon81773980’s problem, because you get a refund after the period is over; and it solves @Blindsite2k’s problem of forgotten/lost accounts eternally holding onto a publicID.
Cons:
This wouldn’t really solve squatting, unless the squatter forgets to renew the publicID (i.e., repay the fee). I’ve also read in other threads that SAFENet has no concept of time, so I’m not sure how the network will know when y is over (though, this is a problem with the regular rental model as well).
Forgive me if this idea has been proposed before in the other threads about DNS; I’ve skimmed through parts of them before, but their lengths are pretty daunting…
