How hard would it be to develop a 'crypto currency only' close-group-consensus network based on SAFE network code?

A couple of years ago, test networks with vaults from home functioned for a number of days, and sometimes weeks at a time.

This showed that the basic concept of the SAFE network was functional, but it wasn’t ready for prime time due to various issues, including spamming of vaults and other challenges of losing consensus if I remember correctly.

Would these tests have failed if they were forced to have a far lighter load?

How difficult would it be to create a cut-down & simplified close-group-consensus network with the sole purpose of running a crypto currency and token system that can avoid being strained to breaking point like the early test networks?

Another way of phrasing this could be:

Would it be much easier / faster to create a secure close-group-consensus network if the requirements of the network were greatly constrained, compared with creating the full SAFE network?

Such a network could have a very light load, as it would only need to:

  • handle currency & token data (vaults would be a few MB or less)
  • handle wallet software (like Decorum Wallet, or Safe Wallet that take very little space)

My thinking, which may be mistaken, is that it would be easier to make a secure / stable network if the requirements on the network were significantly lower than what will be required for the full Safe Network, enabling compromises that boost stability & simplicity at the expense of performance.

I’m not a developer, so would appreciate input on how such a network could be made secure with as little further development as possible, possibly including measures like:

  • Higher number of duplicates of each data piece due to the tiny size of the data being stored (making churn less of an issue, and data loss less likely)
  • Aggressive rate limiting or similar to keep resource requirements low and prevent spam / attacks and make churn a non-issue
  • Built in transaction fees (when required) for all token / currency transactions to prevent spam / attacks
  • Proof of stake node incentivisation (as farming isn’t coded & wouldn’t work if people aren’t paying to host data)
  • Simplified recovery mechanism in the case of network failure that only makes sure token & currency balances are recoverable

These features may be tricky to integrate (some may not be reasonably possible), but I wonder if something along these lines could be far quicker to develop than the full SAFE network due to the greatly reduced requirement on the network in terms of resources.

Could it possibly even be based off the old functional test network code, but with tweaks to constrain that network to avoid the issues that caused its downfall, e.g. by ensuring the network is never anything close to strained?

Why consider this?
For now, I’m more interested in IF and HOW this could be possible rather than why, but if possible it could be valuable because it could;

  1. Show that the consensus mechanism that the SAFE network will use is functional
  2. Boost interest due to having a working (even if limited) product, helping attract interest & devs for the SAFE network
  3. Be an innovative & unique crypto currency and token platform that is fast, has low fees, is anonymous, and has network-hosted wallets
  4. Be used to host MAID, so people have an option other than Omni (though I have no idea about how exchange integration could be achived)

If this would be as hard to achieve with a high degree of stability as finishing the SAFE network, then it’s a non starter, because SAFE will be able to do all of this anyway.

If it could be achieved in a fairly short period of time (e.g. 2-4 months), then it may be worth exploring how it could be funded / staffed, and whether the benefits would likely outweigh the costs.

4 Likes

This sounds awesome let’s try

4 Likes

My first thoughts were to limit the address space of the network to that of MDs used to hold coins (each coin is a MD). The network admin would initially own all coins, and could transfer ownership of some coins to users as they join the network. But as I though about it more, I think other MDs are still needed for user accounts though, and probably also for wallets (although wallet data could just be held offline locally). So there is still some storage requirement, and without mechanisms to pay for the storage, I think you ultimately end up with something looking a lot like Alpha 2…invitation based with put limits per account. Would be interested to hear if anyone else reaches a different conclusion.

2 Likes

This is too risky, coins like Thankscoin or Decorum’s option to create coins are enough imho.

It would be fun to see a lightweight SAFE Network that is crowd runned, more interesting would be if we could experiment with tokens. Think smartcontracts, token generation (ico), maybe even minable coins like Seneca describes it (I’ve secretly been looking for someone to create a coin that you can mine on the SAFE Network, with no luck). Who knows we could even get so bold to create something that works like SAFEcoin by doing a bounty for interested devs. Imho the logic next step is to have a mobile app wallet to play around with these tokens.

:stuck_out_tongue:

4 Likes

second this initiative as alternative to ERC20

1 Like

I suspect that to be secure, then you need all the components for a minimal system (beta) to be at release candidate stage. Once we are at that stage then you have an operating SAFE network. But still not confident for release.

2 Likes

In fact, latest one was launched less than one year ago (test 16 in April 2017). But yes, this seems an eternity from now, I really miss these networks with vaults from home.

4 Likes

I appreciate the time taken to post the OP, but I just don’t think the product is there yet. I was going to post about launching a new community network, but we know that will not handle churn well without data chains.

IMO, there were two big lessons learnt from home vaults. 1, if something is free, someone will consume the lot just to be a dick. 2, we need data chains to handle vaults leaving/joining frequently. I think it probably took home vaults to prove 2, which is why we are waiting on alpha 3, instead of a beta without it… some things you only discover when you try it in the wild.

However… you could already do much of the OP with alpha 2. We already have safe and decorum wallets and can create and move tokens around. They would all be stored within the alpha 2 walled garden with no assurance that they wouldn’t disappear, but it works. I wouldn’t put anything valuable on alpha 2 though.

I think there is someone nice in proving it can be done, perhaps making a bigger song and dance about safe wallet or some such. Maybe even have TestSafeCoin tokens as give aways to challenges on the forum (instead of real safecoin or money). It could be fun that way and maybe they could be used as ‘likes’ of sorts (see decorum).

IMO, if the fun and proof of concept angles can be the focus, it could be a nice way to explore the potential. Personally, putting anything of real value or getting exchanges involved etc would be bad though… It just may all fall apart and alpha 2 is going to be with us for months, rather than years.

7 Likes

I’d love to see it happen and support it if it’s feasible.

Do you think it’s feasible? What changes do you think are needed to make the network secure quickly if it’s scope & capacity is greatly limited?

Great thoughts. Yes, there would need to be storage for account details, balances, and wallet software. But these are tiny in terms of resource requirement on the network. They could also be greatly ‘rate limited’ to prevent them being any burden to the network.

You shouldn’t need mechanisms to pay for storage, other than transaction fee to pay for transactions and possible account detail updates (if this were an issue for spamming the network).

Safecoin wallet Git repository is only 100kb to download, so assuming wallet software is under 1mb, I don’t think that’s a significant overhead for each node to handle without payment, even if they were to host 3 implementations plus test versions.

Nodes would be incentivised to stay online through a rough & ready proof of stake system, where the network runs through a distribution curve & handing out coins according to a schedule, proportionately based on how many coins each user ‘stakes’ in a simple ‘staking wallet’. Details would need to be worked out, but distribution would be more like a traditional crypto, as there wouldn’t be a market for resources like with Safecoin.

Put limits may not be required if rate limiting is significantly harsh to prevent account creation network spam, and transaction fees are in place for any transaction or updating of account details (account creation would ideally be free, even if limited in terms of speed).

That depends on how secure it can be, and what recovery mechanisms are possible. But yes, it would certainly need to be proven to be secure before anyone would trust tokens of value to it, and early experiments would obviously just be experimental before it can be demonstrated to be secure & recoverable.

This might be the case, but if greatly reducing the scope & resource needs can greatly reduce requirements or enable easier fixes that wouldn’t be viable for the SAFE network, it could enable a useful network ahead of SAFE beta etc.

Thanks for your input.

Regarding being free to spam (1), I think that could be easy to address via transaction fees & rate limiting. The scope for spamming at no cost would be very small (possibly only account sign up, which could hopefully be dealt with by rate limiting).

I wonder if point 2 can be fixed on a low-resource network without data chains through a combination of;

  1. vastly reduced resource requirements (churn events are processed in seconds so don’t become a problem)
  2. larger group sizes possible due to low resources per group (churn events affect a smaller portion of any group, so have less impact)
  3. node incentivisation through proof of stake reduces likelihood of churn events, as nodes want to maximise uptime

If it’s feasible, I certainly would as well!

I miss my homey @bluebird he use to whip up community networks within hours. I’m tired of talking let’s do it! Hmmmmm does anybody around here knows how to set up a community network?

The best marketing for the SAFE Network, is if we create a coin that you can mine on it, I think that all Coinhive’s mining power will move over from Monero to it. Maybe even every mining resource, since all miners won’t be mining for blablabla…

I’m even willing to put up a bounty if someone can create a mineable coin.

:stuck_out_tongue:

1 Like
2 Likes

I would see it as a testnet, so any transaction fee or proof of stake would be meaningless. Also, rate limiting account sign-ups would just mean that any attackers could make it impossible for any ‘real participants’ to join

They wouldn’t be meaningless if the coin were required to send transactions, because it would introduce scarcity to prevent spam. How would you spam transactions if you had a limited number of coins required to pay transaction fees?

Proof of stake would be required to mint new coins that are needed to send transactions, so they would have a function & provide scarcity of transactions even if the coins were considered play money.

As I understand it, rate limits would be applied on a per user basis, so they wouldn’t stop anyone signing up, they’d just prevent people from attempting to spam the network by signing up many times at a fast rate.

Saying that, other mechanisms may be required to stop sign up spam if attackers can spin up millions of nodes in short order. If signing up requires only a few kb of traffic & can only be attempted periodically from any node, I expect the scale of an attack would need to be large to strain even a small network.

A “currency only” network would be a yet another shitcoin. Basically you are asking to neuter the SafeNetwork into IOTA.

3 Likes

Except it would be a high privacy, highly scalable crypto currency & token platform that’s built on the Safe network’s consensus mechanism.

Some people have found uses and value in crypto currencies and token platforms.

I don’t know if it could compete with IOTA (which I know almost nothing about).

Though as IOTA has a market cap of $5bn, I’d be pleased if it could be considered on the same level!

1 Like

You are missing the point, and clearly don’t understand where the real value lies in the SafeNetwork.
The very fact that you are happy of a large market cap based on no value at all is alarming.

The real reason that the SafeCoin token would be on its own category is because it is the first token that had an actual intrinsic value based on actual useful resources.
Remove the resources, and you end up with a baseless speculative piece of crap. You are removing the only thing that adds added value to the token.

Having said that, if you wanted to add all those properties (security, resilience, anonymity) you will end up with the SafeNetwork itself once again, technically.

It is akin to asking “wouldn’t it be great to have a bicycle with one wheel?” And after a while, you ask yourself but what if it had more stability on one axis only?.. so the only configuration possible is to go back to the bicycle.

3 Likes

would you really store your coin on such a network? Oo

I do see the motivation and I 100% understand and agree it is worth considering (to some degree) but safecoin is just a special data tag … if you manage to secure the data you basically have the safe network …

And you would need to introduce so many limitations that an attacker might not be able to attack the network effectively but the spam prevention measurements would be active for all users … and we wouldn’t be able to show what makes us different :hushed:

My very private opinion: energy can be invested way better somewhere else …

3 Likes

I’m not suggesting replacing the SAFE network, but getting a limited and yes, less valuable, network together that serves a purpose ahead of the SAFE network being ready.

If you see no value in crypto currencies or token platforms, or less capable proof-of-concepts for the technology behind the SAFE network, feel free to move on and ignore this.