Continuing the discussion from How fast do you think new APPs will be developed after launch?:
Also, adding this:
@fergish It doesn’t matter how minimal the cost is, though. It would create an imbalance in the ecosystem. And if it’s just as easy to download an app that aggregates thousands of accounts into one file system, then why bother with buying the space? We’re talking about about turning 1MB into 10TB. Same way I could create millions of bitcoin addresses on a home computer.
An analogous problem I see: What if Bitcoin addresses all came preloaded with 1 satoshi? That would have been abused in seconds. And where are those satoshis coming from?
Or, what if I create an app that generates an ID on the Safe Network, uploads 1mb of random garbage to its free account, and then repeats this? It would fill the network with garbage, and I’d imagine it would cause a huge problem for the price. Right? Or am I missing something fundamental to the system?
@Blindsite2k, in regards to: “Do you want to give up space on your hard drive to store things on dropbox? No, you want to store a copy of your files on dropbox.” I don’t want a lot of things. But you need to create a balance somewhere, and I think that’s the best way to do it. It’s better than selling my identity to advertisers.
Robo-generating accounts would be a big problem. We’ve talked about this at length in some older threads. Not settled on the mechanism, I don’t think, but general consensus is requiring some sort of CAPCHA in the login process. This would put a barrier that hopefully would defeat robo-login and disincentivize repetitive manual account creation. That was where it was at last I checked.
As far as an app that aggregates a whole bunch of accounts into one file system, I’m pretty sure the network design and dynamics of interface will prohibit this being feasible. Need to check with @dirvine on this. Each account represents an independent “state” with regards to the network, and need to be created by direct handshake with the network. Aggregating multiples would be complex and unwieldy, if possible at all (which I would bet it isn’t). Good test for testnet 2 or 3.
On the other hand, you can take one account and make it look and act like multiple ones, I think.
Bitcoin address are another thing completely–not the same at all.
Yeah, I’d be curious to hear from @dirvine on this. Because I can certainly imagine how this would be absurdly easy.
Could you explain how this is different than generating a bitcoin address? It sounds to me like accounts on Safe would have to be generated using some sort of similar system, since it’s being generated autonomously, it probably has to be cryptographically based and something that could be done on the user’s side (since it’s open source). Just a series of numbers and letters the network recognizes. Or am I totally off?
I’d rather us disincentivize bot-creating accounts with human engineering instead of technology, because it’s only a matter of time before those CAPCHAs are obsolete and the arms race continues. I don’t think giving out a little free space is worth the potential meltdown of the entire network. I sound dramatic, but if I can flood the network with data for free, that’s like, the worst thing ever.
it would be similar to creating lots of dropbox accounts or bitcoin wallets, so possible, but very hard. You would need to build a tool to aggregate those accounts and take bits from different disks and make it look like one. ITs all possible, but very unlikely IMHO, I think even the likes of when google came out with gmail and some folks could use the space as a remote hard drive, Google never bothered to patch that at all.
Can never be sure, but these things tend to sound feasible and may be for a small minority. The hassle would be pretty significant though.
Also remember we have space for 2^512 accounts so that more than the number of atoms in the visible universe. A dead account will only be a couple of integers. An active account now will also take up circa 65bytes so it is probably not a worry from that perspective.
But that’s really easy! I just wrote an app that generated 100,000 bitcoin address just now, while I was making coffee (this is a lie).
And the only annoyance for Dropbox is the email address setup. But Safe doesn’t have anything like that. Not to mention Dropbox has people monitoring that, and can have a human deal with it, and track IPs. Safe, being autonomous, just isn’t capable of monitoring that.
Maybe I’m unclear on the account creation process?
You might be misunderstanding my concern. If I created an account, and I were to get 1gb of free space from the network, what’s stopping me from coding a bot to make millions of accounts, then all of those accounts upload 1gb to the network. Then add a botnet to that and I’d imagine you could really damage the network. It’s not about all the accounts, it’s about the size of the space available to them for free.
Basically, what I’m getting at is creating a bot that consumes resources across the network without compensating anyone for them.
PS - I swear I’m not trying to be a nagging baby! I just don’t see what makes this difficult yet.
What about rewarding the free amount only to farmers?
Surely, it must be harder to maintain multiple vaults in good standing (higher than rank 0) compared to multiple bot accounts. This would require actual resource dedication: electricity, online time, bandwith, hard drive space, data caching, etc.
If farmers earn (free storage) they are able to spend their Safecoin on other things like APPS, SAFE Network stores, exchange for other currencies, etc.
Right now, GET requests are free, so a user can create an account and “browse” the Network. They can become a farmer to earn some free storage. If they don’t want to become a farmer they will have pay Safecoin to buy storage. Hopefully, it will be very cheap.
I think the demand for storage will increase exponentially because of the APPS. Network “interaction” requires PUT requests which consumes storage space on a crazy scale. I expect usage to increase as more APPS are added to the Network.
I’m sorry David but what you’re saying doesn’t make sense due to unclear terminology. I really don’t mean to be didactic, but I’ve really been trying to get a point across about this for a while now, so Hark! (Exacting explaination warning.)
Your statement above is technically identical to saying, “What about rewarding the free amount only to persons who pay with safecoin which they’ve earned by running a farming vault.”
A person can own and run a vault that handles network business and farms for safecoin. That PERSON is then called a farmer.
A person can enter into a relationship with the network as a user. That PERSON is then called a user.
But the network doesn’t recognize that person, only vaults which may be farming, and users. The functions represented by the terms farmer and user have no direct network association, except that the farming vault is told to send earned safecoin to a specified user account. But the user account the coin is sent to doesn’t have to be the same person who operates the farming vault, so even where the earned safecoins are sent doesn’t make that connection.
The reward for a farmer is safecoin. Period. The person who receives that safecoin as a user can spend it on the network for more storage or whatever, just like a user who buys or is given safecoin.
If there is a free storage amount, it will need to apply to all accounts. There is no inherent network association between a “farmer” and a “user.”
See what I mean?
I’m starting to think that @chrisfostertv may be right: Allow a new account enough free space to be able to browse and keep bookmarks, store some very small amount of such personal data, and buy more storage with safecoin, however acquired–farmed, bought, donated, etc. This free level would probably be in the Kb range or the low Mbs.
This would make rock-and-roll adoption a bit more difficult, but would barrier the likely front-end attack of creating lots of accounts and jamming the network storage resources up to the eyeballs with junk.
Now, I wouldn’t be surprised if @dirvine, that clever fellow, hasn’t thought this through and have it all figured out. But that’s how it’s looking now.
If not for that attack vector, think at least a few gig of stoage would be great, is not more.
don’t think that was me, it sounded brainy
That’s right actually. I think it was @janitor? (Not that you don’t sound brainy now and then. )
I don’t know if you’re trying to argue my use of the word “farmer” or if we cannot make the Network do what I’m suggesting. I’ll try to lay it out.
Can the Network also award free storage to a user who is running a vault? I believe so.
- Every vault must have a Safecoin wallet address.
- Every wallet address belongs to a unique user.
The Network could award storage credit in the same way it awards Safecoin.
The Safecoin Wallet
- keeps a balance of your Safecoin
- keeps a balance of your NSL
Similar to GET requests that trigger a farm attempt to receive Safecoin, vaults get upgraded in rank based on good behavior, evaluated by the Network.
I am suggesting when a vault is upgraded from rank 0 to rank 1, to rank 2, and so on… a free storage amount (bonus if you will) be credited to their wallet. The higher their rank goes, the more storage is credited. We can qualify the credit amount in various ways: available_space, or stored_space, online time, etc.
I know the current protocol is not coded this way. But if this idea solves multiple bot account abuse, then why not? I’m trying to innovate with new ideas, solutions. If the Devs think I’m wasting their time, I’ll be happy to get out of the way.
Yes, thank you for the reference.
I made a comment about that here.
Make account creation cost the amount of safecoin that storing such an account costs. This is the only true solution to this problem.
This way anyone can give out “free” accounts in exchange for say a 6 month old twitter or facebook account. Or just pay a few dollars for the safecoin directly to be able to acces it without associating social network accounts with maidsafe ones. This way the crug would be external to maidsafe and easy to manage, while maidsafe only concerns with running the service and charginf the market rate for accounts.
Maybe I don’t understand you, or maybe you don’t understand how MaidSafe is supposed to work.
- You buy “SAFE” or Safecoins on the open market.
- To open a SAFE network account you don’t need anyone’s permission or help. You just get the s/w and install it.
- It’s not required to associate your MaidSAFE account with any external information such as a social network account or email address
Is this how you understand it?
I agree this would be best, initially the issue will be distributing enough safecoin for folk to jump start the network though. If there were enough farmers and some clever marketing/seed rounds then perhaps we can get momentum. I will write about this sometime, but there are a few pretty neat options for us.
A disk can be had for $0.05/GB.
Clearly noone who has a computer or mobile phone should have any problem paying for the first GB or five.
While my position on freebies is known, here’s couple of ideas for you:
- X Safecoins for MaidSafe newsletter subscription
- Y Safecoins for MaidSafe newsletter subscription by a validated IT person from an NGO (1 per NGO)
- Z Safecoins for top 3 MaidSafe apps submitted by individual developers before end of 2014 based on forum member votes at maidsafe.org
- T Safecoins for each validated exploit submitted directly to MaidSafe Development
Then why not just use the safecoin? It’s already there. It’s already earned. There’s no reason for an extra protocol function or layer. Running a node will earn enough safecoin to buy much more than a comparible amount of storage.