The massive deflation of year 2025, rise in value of Safecoin, how to cope with it?

Division is definitely preferable. You can divide for infinity, to suit market requirements. If it can be done effectively, it has to be the best option.

3 Likes

Agree with Traction. Also it should be simple and as user-friendly as possible. Mainstream and common people do not like complicated things. To use other Altcoins for Safecoin change is too complicated. Maidsafe network is designed for common people and should be simple and understandable to be adapted by mainstream.

3 Likes

Yeah what’s so hard about dividing?

You say it’s hard because it’s not a ledger, but why not just let people send fractions of a SafeCoin to each other, but only full SafeCoins to the network (for PUTs)

What’s hard about that?

Does it stress the network or something, which would make zero tx fees impossible? I just want to know

On safe net, every coin is an item of data. So, it is a question of efficiency.

If it is decided to just issue lots of safe pennies instead (say), then transferring one full safe coin would require 100 safe pennies to be sent.

However, if we enable splitting and joining, we can be more clever. If we want to send 10 safecoin but only have 1000 safe pennies, 2 operations could be performed - one to convert 1000 safe pennies into 10 safe coin, then another to send the 10 safe coin. The reverse could be true too, so that we split 1 safe coin into 100 safe pennies, if we only need to spend 10 safe pennies.

With physical money, it is easier to carry and exchange large sums with notes - they are less heavy/bulky. However, sometimes you need smaller denominations for small purchases. We just switch between these naturally to suit. This will be the same for safe network - dealing in pennies for large payments would be inefficient, as there would be lots of passing pennies about and counting going on.

This is also why just having safe pennies alone will be inefficient. Having to pass thousands of pennies about, just to pay someone 20 safe coin would seem undesirable.

Ultimately, we just don’t know how fast or inefficient either option is going to be on the real network until we do some benchmarking on structured data ownership changes. Even if shuffling safe pennies is actually relatively low cost, we would have to be certain that no smaller denominations would be needed.

7 Likes

My fear about “costless” transactions that break a safecoin into 100 or 1000 microsafe is the ability to spam/DDOS whatever node/exchange performs such a transaction. Without a “stake” at risk in the transaction what stops someone from using resources converting back and forth millions of times making change or grouping change?

Presently some type of POW (memory hard/GPU hard) needs to be added to Launcher to stop login DDOS possibility or at least make it impractical. Possibly solvable by having the login node(s) ratchet work requirement up if request get excessive in general or to specific “users” (multiple attempts in N time) . Be it Scrypt combined with a certain amount of PBKDF2 iterations according to network congestion in that space, or something else, its solvable.

“SafePennies” without Xchange fee would create a huge DDOS potential on any exchange or internal service running it.

Division or expansion should work fine if built in or made to happen after a certain time frame.

2 Likes

That makes sense. Crazy tx spam

Just use alt-currencies for smaller denominations then I guess

Well the method of division will affect how much a problem this is. If you make every “micro”/“milli” SAFE into an SD like SAFEcoin is an SD then I would agree. Easy to keep on sending the minimum amount out and then have another send it back.

Although like SAFEcoin if one tries too many transactions too quickly then the groups handling it will begin to fail the transactions.

So spamming will not have the same effect as it does on block chain.

Then there are other options to divisibility that will not fall foul of spamming so quickly.[quote=“bellyfat, post:45, topic:11444”]
Presently some type of POW (memory hard/GPU hard) needs to be added to Launcher to stop login DDOS possibility or at least make it impractical
[/quote]

Then those wishing to abuse it will rewrite the launcher to remove it. At this time it is expected that a login sequence will take time, around a second to return the account ifo. So it is limited how many login attempts one can do.

1 Like

Even failing transactions puts work on the node/groups doing the “failing”. The disincentives should always if possible be visited on the attacker. Maybe my understanding of the topology is off but if I am the one attacking from say 10,000-100,000 accounts (virtual instances or botnet) it would seem to be the “fail” list/filter would grow rapidly (process time to lookup) while I (attacker) just reduce the attack cycle time till these transactions pass the window, or on board more bots/instances.

Some hardening in login, account creation will be needed quickly. The division issue doesn’t create an attack vector at all if its SD just like regular Safecoin, but as an alt or an exchange I see a strong attack vector.

Having a node/group get the power of a traffic cop by increasing costs on any launcher in its space/group with this “ratchet/work” can help a group mitigate attacks and congestion, slow login DDOS potential. I strongly believe these problems should be solved before Beta launch.

just realized, whats stopping ppl from sending regular safecoin to themselves over and over forever to try to harm the network?

why is this only an issue with divisible safecoin?

1 Like

Perhaps just charging a small fee to split coins would suffice? If it is a cost to the network, it should probably be reflected. It only needs to be big enough to prevent abuse.

The node should be able to “fail transactions” that are going back and forth between 2 wallets, but maybe not 6 wallets since it could appear “new” on 4th user coin send. But if the group is smart enough to know how many are in its close group a certain amount of sends of coin should for those “accounts” possibly cause those spends/sends to fail it seems to have been implied. My suggestion was just in putting some of the costs to the launcher/user if that groups traffic is excessive (spammy). Keeps the spends “free” if its POW in the launcher app that responds to the groups traffic signaling. Has nothing to do with “mining” but only to do with rate-limiting abusive account creation or sends/spends. It seems with regular “safecoin” you cannot easily target a group to attack so sends would be “diffuse” due to network topology. Still requiring work of potential bad actors in the launch client could eliminate this as an attack vector altogether. No appearance of spammy acting clients in group, POW would never be requested.

sounds bad, limits

makes me worry honest people will have trouble sending safecoin at some point with these blocking limits

In the actual implementation exist a quite hard Scrypt password hashing.

About the safecoin spam, the idea is that the computational work for the sender will be high enough to be inefficient as attack. Remember that, unlike bitcoin or other cryptocurrency, the sender must sign and send the order of each safecoin or each division. If the attacker wants to really bother the network with thousands or millions of transaction must made an enormous computational work.

6 Likes

Already “in there” like Ragu sauce. Thank you for the reference.

1 Like

@mvanzyl shared this back in march.

“The Safecoin Whitepaper states each coin is divisible something like 4.3 billion times.
I think they have addressed the issues very well there.”
The link to that PDF is 404. Not sure it this is still accurate. If it is then there will be a possible total of 4.3 billion squared total coins or 18,490,000,000,000,000,000 units. So thats something like 18.49 quintillion units where 2^32 units equals 1 safecoin. This would allow for sub nano payments even if 1 safecoin were 1,850 usd each. wow

1 Like

wow, safecoin is the best thing ever

free, spam-proof, 100% annonymous and the fastest confirming cryptocurrency ever created.

this is crazy. People will FREAK when TestSafeCoin is available to use and try out :slight_smile:

gonna have to get used to super cutting edge currency like btc looking like caveman drawings on stone :slight_smile:

2 Likes

I don’t think that is a set goal at the moment. More of what can be done since there is 64 bits. The issue with creating upto 2^64 SDs to handle division is what @bellyfat hints at, the transaction load if one is to send 1/2 a SAFEcoin worth, that would be 2 billion transactions if the coin was divided that much.

There are other ways and I am thinking in terms of division upto 10^18 times. But also to help mitigate issues with people sending the smallest amount over and over, is to have a “minimum transaction amount” which the network can handle taking into account

And not have individual SDs for each division part of a coin. And the minimum transaction amount is set in code which can be reduced on a new version to reflect the greater network size.

3 Likes

irvine did state before he had ideas to create 3 to four different splits for different amounts iirc.

I’m not sure farming would work with that many coins. Remember, awarding a coin depends on if its address is already in use. If it’s virtually never used (which would be the case with 64 bits) then “all” farming attempts would result in getting a coin, and that’s not how it’s designed to work.

1 Like

Very true and very important. Without it there is no reducing rewards when the coins available is low.

I have thought of a solution which is part of my idea coin division/wallets discussions coming soon

1 Like