Block 752 here.

Yeah, this part isn't actually set in stone, so if other people have input on this, don't be shy! Had a discussion with Cygnus and Cujo about this last night.

I'm thinking the best path might be a compromise between people wanting a set-limit system and an actual currency's behavior. Bitcoin and almost every other cryptocurrency has some finite limit to it's mintage, attractive to investors as it is scarce, similar to a commodity. On the other side, fiat currencies are designed in such a way that their mintage is based on a lot of factors including population, and tend to become less valuable over time, per unit. As a result, a compromise would be a very long mintage schedule with an extremely gradual decline, so the coin is still generated at a predictable rate, but is still limited in supply in a given period of time.

Hey Indigo, good questions!

1.) Yup, full support for multisig addresses/transactions will be available. The process behind it will likely have to change due to the signature scheme we are working on, so there may be multisig 'accounts,' in which the blockchain has a written record of all the addresses involved in a multisig account, and can apply appropriate verification rules to make it behave similar to multisig. This will also allow you to sign on additional parties to the account.

2.) Curecoin will have a new API, since so much under the hood is changing. While we don't currently have a RPC reference of any sort, we're rebuilding the structure of interfacing with the daemon from scratch, so while some RPC calls will be very similar in function, there will be far more complicated ones to account for additional features.

3.) The blockchain should be secure enough using only a certificate system, given that certain rate-limiting rules are integrated into every node. However, I'm not opposed to putting some kind of Proof of Stake or Proof of Capacity into the currency. If you would like to look at a current example of Proof-Of-Capacity, check out Burstcoin.

4.) There will be a limited amount of coins that can be minted using certificate mining (however this will be stretched out over a very long period of time, and will be practical for encouraging computational research far into the future). If we implement a Proof-Of-Stake model, then we will see some form of interest, which will be rate-limited, but not limited to an actual final quantity. Hopefully that makes sense, let me know if you'd like me to clarify that one.

5.) Unfortunately the design of a certificate system would require some authority to issue the certificates. For a very specific set of mathematical research, the solutions could be proven in the blockchain, however we don't plan to implement that feature, unless we can find a robust way to do so. Our logic essentially follows that *most* ground-breaking research requires fairly significant human intervention, from creating the workunits to be crunched for protein folding to developing climate predictions based on variables that scientists want to manipulate to best study their subject area. The few exceptions to such research are of course ones like Primecoin or Riecoin where the network looks for numbers that fit a given pattern that can be verified by the network. In order to do so, the Curecoin network would need a means of validating the work submitted, which, given the scope of potential research offered by deterministic work generation, doesn't seem terribly practical. That being said, if an extremely interesting research project which could behave in a decentralized manner was developed, the validation of that project could be added in a similar manner as adding an additional university to the network--and would require a client update.

I'd like to say it'll be fully released by February (full crossplatform GUI, crossplatform daemon, exchanges, cc1.0<->2.0 trade working, etc.), and we'll have public (no value) betas before that, but you know how software development goes. I also don't know how much time it'll take to get universities on-board and signing certificates, but if the community is fine with it, we could be the issuing authority until the universities are ready to switch over. I'm not a huge fan of that idea, but it's on the table.

Moving right along!

Ah good point, I'll fix that.


Similar idea, however we are completely unaffiliated.

Took a break from coding curecoin core to do a bit of preliminary GUI mockups. This is what I came up with:

You're probably looking at somewhere between 4 and 5 days, depending on how good of a mobile i7 you have.

Not that I know of. Here's some basic performance metrics on Centos 6.5 with dcct's linux plotgen:
i5-2400: 3400 nonces/min
E5-2560v2: 7000 nonces/min
E5-2670v2: 7100 nonces/min
2xE5-2560v2: 13900 nonces/min
i7-3770k: 4300 nonces/min
FX-8350 @ 4.2GHz: 3500 nonces/min

Oh, we were never on Mintpal. Just don't want to get added and then delisted due to low volume.

For solo or pool mining, I would need to generate the plots for your own address, so unfortunately I can't pre-generate plots. However, I can deploy storage very quickly in the cloud.

I have 4TB drives available to start generating plots immediately (which would take 2-3 days to complete) after which I could ship the drive. For other capacities, I would have to wait a few days for shipping to me first, so a 6TB drive would likely ship from me in about a week. The included shipping is UPS ground, though if you wanted the drive faster I can offer faster shipping for an additional fee.

Hey guys, thanks for the interest in MintPal! I'm afraid Curecoin is too low-volume right now to float on MintPal unfortunately, and I doubt MintPal would be likely to re-add a coin that they already delisted due to low trading volume.

If enough people are interested, I'll reach out to MintPal, but I was personally planning to wait until the volume began to pick up a bit more, maybe closer to cc2.0 release, when we have some solid pieces of infrastructure proven working, and there is more community interest.

For anyone interested in purchasing pre-plotted external hard drives, or renting cloud mining for USD/BTC/BURST:

http://www.vorksholkms.com/

You can check out OVH's dedicated storage servers, which you can grab 48TB for $489.00/mo. That's around $120/TB/year. Or if you wanna go biger, 216TB will be $1659.00/mo, or around $92/TB/year. Keep in mind you'll need to generate all the plots too, so you might want to vRack that with a powerful "Big-Data" server such as the HG (which could generate plots at a rate around 13900 nonces/min, as it has 2xE5-2650v2) or an Infrastructure server such as the EG-128 (which you can rent weekly for $100, and could generate around a rate of 7000 nonces/min) for a few weeks to generate all the plots, constantly moving them over the vRack connection.

Alternately, you could rent SYS servers such as their 24TB option for around $150 per month, or around $75/TB/year. Keep in mind that every TB of stated space is 0.9TB of space after formatting and creating the filesystem, so the 24TB server will give you 3.6TBx6=21.6TB after formatting and using "tune2fs -m 0 /dev/sdx." I also rent out burst mining power at a rate of 0.045BTC/TB for the first month (for an actual full Terabyte, or a Terabyte *after* formatting), and 0.0225BTC/TB each month afterwards. If you use SYS for the servers, you can still use OVH Enterprise Dedicated Boxes, and then download the plots over the internet. However, this connection will be limited to 500Mbps, and in actuality you're performance will waver between 300 and 400 if you are saturating multiple connections constantly.

Interestingly enough, Hetzner.de offers a 16TB dedicated server for 69 Eur including 19% VAT (so around $74/mo if you don't live in Germany, so you don't have to pay VAT). The processor in that box is capable of somewhere in the ballpark of 3600 nonces/minute more than likely, which would give you just over 1.18TB/day of plots. 16TB after formatting is 14.4TB-ish, so that would take the server around 14.4/1.18=12.2 days to create all the plots. Renting another dedicated server for plot generation seems impractical with Hetzner given the 200Mbps/connection, which may or may not also be the download speed.

I believe someone did a preliminary test and got the equivalent of 35GB per 4-minute period, so effectively 35GB per block on a high-end GPU, which drew somewhere around 200+ W at the wall.

Already providing that (0.045BTC/TB for setup + one month, then 0.0225BTC/TB a month after), but some people would probably prefer to outright own the hardware if they're looking for long-term BURST mining.

Would people be interested in a service where you can buy pre-plotted external hard drives with a one-click miner?

Sent my promised 25k to your address in sig from BURST-AAAA-AAWN-27RC-64PAA.

It's significantly faster, and will work on any Linux system which uses makefiles and gcc--Opensuse included.


You can edit the Makefile to add -O3 and -march=native, though I've seen no performance boost from doing so personally. On a 2xIntel E5-2560v2 (so 16 cores, 32 threads) this plot generator gets around 14000 nonces/minute, while the traditional linux plotter gets a bit over 7000.