20 replies to this topic

[RealSolid]

Well actually it restarted about a week ago but I've had the flu for 2 weeks so didn't feel like talking much!

There is some rather big news coming out soon regarding new features. I'll divulge one of these for now.

We are removing the concept of "blocks", you can now mine MicroCash in a manner that treasuries around the world print their money. This also means there are no wasted shares or competition for limited amounts of blocks. You select the denomination and mine away until you find a solution for that particular value. If if you want 1MC, you can select that, if you want a 100 MC "note" you can mine that. You then submit the valid cash to the network, or you can choose not to and just keep it "Hidden". It's an interesting concept that we can have "cash" that the network doesn't know about but is 100% valid. It could mean actual physical notes could be printed that have nothing to do with the live MicroCash network but can be redeemed on it at any time by any one. Much like how 'real cash' isn't in bank accounts.

The BETA is getting close to being ready, I'll know exactly how much longer when the rest of the features are announced in the coming days.

[rlh]

Wow, I really love this idea and am quite interested in the technical details of such an implementation. I know you will reveal the details later, but how will the PoW work for these mined notes? I really love the idea. I'll have to find a way to implement this into a MLAF project.

Thanks for innovating. This sounds really cool.

[viperjbm]

I am really curious to this, please share more and open up public "scrutiny" and "critique" and "praise" :-)

[RealSolid]

Well I'll also add that each thread of the MicroCash transaction engine gets 10% of Visa's current transaction volume per second. 10 threads = same as what visa is doing right now, in consumer software on consumer hardware.

[CaptChadd]

This is good news and I am so looking forward to giving the beta a go.

[platorin]

RealSolid, what about the coins already mined? We all know they are going to be implemented into MC network with 1:10 migration - fair enough for everyone I think, but what "notes" values will those money have? Can one set how many 1 MC notes or 100 MC notes does he want during the migration process?

[RealSolid]

MicroCash Notes only have relevance until they are cashed. Much like a bank balance has no concept of notes, the live MC network has no concept of notes either. Only people creating currency have a concept of what term of "bond" or "note" they are attempting to make.

So everyones SC will just be given to them like a bank balance transfer.

[rlh]

Since these "notes" can be mined externally from the system, does the demurrage/interest system somehow affect them? In other words, If I were to mine one "note" today, but not add it to the system, would it be usable in 10, 20 or 50 years?

[RealSolid]

rlh, on 28 August 2012 - 12:14 PM, said:

Since these "notes" can be mined externally from the system, does the demurrage/interest system somehow affect them? In other words, If I were to mine one "note" today, but not add it to the system, would it be usable in 10, 20 or 50 years?

Not unless you mined it "extra hard". The reason for this is potential system upgrades in regards to power consumption per hash.

So say you made a 100MC note but it was in fact worth 200 MC in regards to its proof of work. That would mean you could survive 2-3 years at least holding it without cashing in, unless something drastic changed on the mining scene (asics).

The general idea right now is some way to update the mining needed for 1MC for relevant hashes at a most of every 3 months. So 4 times a year max, a number can be changed (or not) to reflect the current mining environment.

If there was no limit then it's possible to abuse it in the future by saying you made a hash in 2012 when in fact you made it in 2015 with much better hardware.

[rlh]

To be clear, the actual value of the mined note will be determined by the difficulty/hashing power of the network at the time the PoW note is posted to the network? Interesting. This is less of a note but more of a bond with the potential for having a "negative gain" in interest.

I would recommend changing the semantics, though. I wouldn't say someone mined "1MC", if in a month it will be worth .75MC when the note is submitted. Instead, maybe we could come up with some other name for this denomination. Something like Proof-of-Work-Units, or PWUs. The value of the PWU note will never change, however, it's value to the system will vary depending on the time that it is submitted. This makes PWUs somewhat of a commodity and... I see it now. A bond market could form around this mechanism.

I still like the idea, even though the value of the note will trend a net loss with time. This means users like myself who are behind proxies during the day can mine on local desktop machines that I can't connect up to the network. Nice!

[viperjbm]

What is securing this system if the miners can mine offline? What is described sounds merely a distribution plan (which is important too, but...)

[RealSolid]

viperjbm, on 29 August 2012 - 10:27 AM, said:

What is securing this system if the miners can mine offline? What is described sounds merely a distribution plan (which is important too, but...)

Mining will no longer have anything to do with "Security", only with economy. And MC will end up supporting multiple hash types but the basic premise is the same, regardless of hash type we will generate 1MC when someone has used about 0.80 in electricity.

In regards to security we looked at multiple e-currencies and see where they have succeeded and where they failed.

Liberty Reserve for instance has more daily volume than bitcoin with 90+% of it being used for services and goods trade. Liberty Reserve payments are instant and have stable value, yet it is a closed system controlled by one company. They can shut down your account, freeze your funds and no one will ever know it because it's entirely private. They also support chargebacks. You also have to go from fiat to liberty reserve, there is no public or anonymous currency creation. A lot of this also applies to Dwolla.

Bitcoin is apparently "decentralized" but 3 or 4 pools currently control the majority of it, with one company (BFL) coming out with nearly 80-90% of control if they decide to use it. So they suffer from multiple attack vectors, 51% attacks, the ability for pools to reject your transactions (as is happening right now with Satoshi Dice being blocked by some pools), a poor economic aspect and transactions which take a long time to confirm. Yet Bitcoin does offer some good things like relatively anonymous transfers and provided it isn't attacked, it's pretty good at not having chargebacks.

So MicroCash is going to use the best of both worlds. It's going to have a p2p network (which makes it much more difficult to take down), distributed currency creation, instant transactions, no 51% attacks, no chargebacks, no frozen accounts and an entirely public record of it all. It will do this by having a gateway node. A node (eventually a group of nodes) which is the same as any other except they are restrict the flow of transactions through a gate, one at a time, so they can be processed in order. This way there is no network subjectivity on what is what, the gateway node tells all the other nodes. Even if the gateway node key is compromised the worst that can happen is network "Downtime".

One thing I realized about paypal, libertyreserve, dwolla and even bitcoin is people are generally forgiving of downtime as long as it's relatively rare. Given this I designed a system whereby the gateway node(s) have no power at all except to restrict the flow of transactions in a serial fashion so everyone can agree on a record. Every node still verifies every transaction and if they detect the gateway node as doing anything abnormal they will immediately go into "quiet" status and wait an update. This means there is no lost funds, no chargebacks, no rewinds and no way for the gateway to selectively reject transactions. Essentially it's 100% secure outside of cryptographic attacks. Any compromise of gateway nodes simply means "transaction downtime", which would take about 5 minutes to fix and be up and running. For what its worth for the year SC2 has been running there have been no key compromises.

[RealSolid]

Another thing I will mention. We were going to support "2 key authing" with transactions for a while now, which meant we needed to store an extra key for each account. I decided instead of just limiting to another private key, I'd let people set that in transactions to whatever they want. So this means there is 64bytes per account that can be set to whatever that person wants. This also opens up potential use for MicroCash to something like shortening web addresses, a DNS system like Namecoin down to even a distributed cloud storage system (linking accounts to get >64 byte storage). This adds no more bloat to MC itself because all these other things can just be "added" on by others.

This only works due to account fees restricting spam and cleansing the system. ie If you wanted to run a distributed web server over MicroCash you're going to have to pay for it in fees each day which pays everyone else in the form of interest.

[notyep]

"Innovation at work"!

[RealSolid]

Quote

D:\projects\microcash>microcash gettx 0
{
"number" : 0,
"type" : "Currency Create",
"amount" : 10000,
"to" : "micro(QSM3PKS3EOZBUZO4A)cash",
"hash" : "FBB87C5E4040B7A7A95999E5FA5BA4BD25AC360A9B3805FCD1317A6997674ED9",
"runninghash" : "44B10CB8182936B29E78F8BF9AD06C7C63EA054C2C784A478115A2794E16578E",
"powtype" : "Solid256",
"powhash" : "00066DE56CD6FD50D2FC187E3639DCA8A2EA6F5C68D9C8135F5F765E03D4DC83",
"info" : "0001000030303031414e4f4e43415348de3c0000000000000000000000000000000000000000000000000000000000000000000000
0000000000000000000000"
}

I currently have transaction mining done and relatively well tested. The next step is normal transactions and p2p testing. I have done some quite extensive changes to how the p2p messaging system works.

In SC and Bitcoin nodes communicate a list of "inventory" among each other. If you get a new transaction or block you then tell other nodes "hey node, I have this new transaction or block which has this hash". It is then up to those nodes to request the block or transaction from you if they don't have it.

Old way
1) Node receives a new block or transaction
2) Node sends out a hash to all nodes it is connected to alerting them to this new transaction or block by giving them its hash
3) If a node doesn't have this block or transaction it then requests it from the node

New Way
1) Node receives new transaction
2) Node sends transaction to everyone except the node that sent it

So we have cut out a step and reduced latency. This also reduces the attack surface for each node as it is impossible to send a fake transaction without nodes banning you. As soon as a node sends a fake transaction it is banned for an hour. Whereas in the old way with a node sending fake inventory it is hard to validate. It's hard to disprove a mere hash as invalid. Due to our small and fixed transaction size it is possible to send the whole transaction at once and it is a better way to handle new transactions on the network. It ensures transactions will propagate the network at close to the speed of light. It is very similar to a "viral infection" except it's part of the network design and everyone has to pay a small fee for the privilege of sending a transaction at light speed.

[Dacm4n]

Looking forward to the beta.

[SuperTramp]

Any updates?

[RealSolid]

I've been working on the networking side and that's nearly finished. Once that's done a beta can be released.

[viperjbm]

sweet deal, good to hear and looking forward to it!

[CaptChadd]

Hi Realsolid, I have 106 solidcoins in my wallet but with no connection I cannot send them anywhere or trade them? Is there a way to get them converted to Bitcoins or something? I still am going to back Microcash aswell.