Topic: [ANN] Sia - Decentralized Storage - page 488. (Read 1382189 times)

Anyone can teach me, how to compile this source code to sgminer? I'm noob   

We will try to make it as easy for the users and developers as possible. We're planning on implementing a plugin system for the UI, you could put in a plugin that limits your host choices to hosts that you are legally allowed to choose.


https://github.com/NebulousLabs/Sia-GPU-Miner

Here is the source code. As far as we are aware, it works on all cards that support open-cl. The mining will not work with the master branch of Sia until the 'hardfork' branch has been merged. The hardfork branch is here: https://github.com/NebulousLabs/Sia/tree/hardfork

We are expecting to merge the hardfork branch either tomorrow or Tuesday.

You will need to make it easy, like "select from a list of countries you permit your data to be stored in".  Users can't be bothered sifting through lists of IP addresses.


The law is the law and companies are required to follow it, whether or not it makes sense technically.  That's just an unpleasant part of the cloud storage market and Sia must allow companies to comply with the law if it's to be successful in entering this market at all.

Right now we just have 1 opencl miner that works for both types of cards. We'll be publishing the source code in the next 2-3 days.


These numbers are discussed at length in post #129. Thanks to payment channels + contract channels, a near-infinite volume can be handled off-chain. There are limitations to the total number of users, however.



http://www.marketsandmarkets.com/PressReleases/cloud-storage.asp

http://www.forbes.com/sites/louiscolumbus/2015/01/24/roundup-of-cloud-computing-forecasts-and-market-estimates-2015/

http://www.prnewswire.com/news-releases/cloud-storage-market-worth-468-billion-by-2018-182524001.html

The numbers aren't terribly consistent, but they are in the $10s of billions per year, which is more than 200,000 BTC per day. It's not actualy Bitcoins being spent, I converted USD to Bitcoins when making the point.



Users/apps have full control over which hosts they select. Though I personally recommend against staying inside of one country (it's definitely less secure), if national regulations force you to stay inside of a country than it's a rule that can easily by followed by checking a host's IP address before agreeing to upload to them. Client-side encryption makes this regulation unnecessary, but governments aren't always savvy enough to realize this.

Thanks for the good answer. What kind of tx volume would equate for 10,000 BTC per day on storage?  Currently how many tx can fit in a Sia block and what is the block time?


Please provide a reference for this.


great.

I do have one major concern regarding the uptake of Sia by businesses.  Many countries have data protection laws that mandate that users personal information may not leave the country.  Sia won't be used by any serious businesses unless it can comply with these laws.  How do you plan to deal with this?

that sounds great ...

will the miner itself be separated into an nvidia miner and amd miner? ...

im asking purely for the pre-setup that needs to be done an this end ... especially if you have the source code available and we can compile the miner readily - from github or similar ...

tanx again ...

#crysx

This is correct.


Yes, you are. Siafunds do not take any fees from miners, do not take transaction fees, do not take subsidy fees. The only fees that siafunds take are from siacoins spent in file contracts. This is very different from transaction fees. If people are spending 17BTC per day on storage, each siafund will be earning 0.000066 BTC per day. If people are spending 10,000 BTC per day on storage, then each siafund will be earning 0.039 BTC per day, or 14BTC per year.

In today's space, a total of around 200,000 BTC are spent per day on cloud storage. If we owned that entire market, each siafund would be earning 0.78BTC per day, or 280 BTC per year. That's a significant ROI for an instrument that cost less than 1BTC in the first place to purchase. Sia plans to replace all cloud storage, and that's why siafunds are valuable.


Correct, the siafund fee applies only to storage contracts. Traditional mining, transaction fees, and traditional transactions all have no fee.


We hope to have a guide out for the various ways to make money. You can run a POW miner and a storage host from the same machine, at the same time. They don't interfere with eachother and each provide a separate form of income. We finished the rough draft of the GPU miner today - it works! We'll be optimizing it and releasing it on the 7th.

Siafund is calculate based on the storage trade/charge/contracts volume. Transactions in Sia are supposed to be free if I am correct.

Just in case anyone wants to verify, Bitcoin is currently getting ~16.8 BTC tx fees per day https://blockchain.info/stats

For the sake of comparison, we cannot use the transaction volume (amount of money moved) nor the Bitcoins mined.

It's only the tx fees that siafund holders collect.

I would really appreciate some clarification from the Sia developers on what kind of tx volumes they expect.  Also what kind of fees are there?

No.  That is trade volume (amount of BTC changing hands) not the amount of transaction fees collected by miners.

To calculate ROI you need to look at the amount of TX fees collected instead.


Really? How many tx can fit in a Siacoin block and how many blocks per day?  I think you'll find it's not enough to get anywhere near "high volume".

Bitcoins 24 hours volume is $14,097,800 - http://coinmarketcap.com/currencies/. If Sia has same volume, one Siafund will earn $14,097,800*0.039/10000=$54.9 per day. If Sia becomes dominant in storage industry, it's volume should be higher than that.

Coinmarketcap.com is a good place to do a quick ads campaign - http://coinmarketcap.com/currencies/

Everybody?

Sia seems to have an astronomically high block reward that continuously inflates the siacoin supply forever.
With an indefinite 30,000 siacoin block reward, how can siafund holders capture any value from their measly 0.000039% from each storage contract?

This doesn't seem "damn good"

Since the NXT AE investors get access to 10% of all 10,000 siafunds, then all the NXT siafund holders combined will get only .39% of the tx fees for storage contracts.

This doesn't seem "damn good" either.

Let's run a quick calculation based on Bitcoin transaction volumes.

Per day (well today) Bitcoin tx fees amount to about 17 BTC.  a meager 0.12 BTC per block

Assuming Sia will have as much transaction volume as Bitcoin (which it won't), then traders on the NXT AE paying $1,000 per siafund could earn at best 0.000039% x 12 BTC = 0.0000468 BTC/day = $0.010998/day?  That would be ROI in about 249 years!  

Even with the full 3.9% Nebulous would only earn 0.039*12BTC/day = 0.46 BTC/day which isn't enough for a single person to live on, def not a company of 1000 employees.

I still think the idea of decentralized cloud storage is good, but I don't Sia-how-da economics of Sia works out good for anyone.

Am I missing some crucial information here?

No. There are 10,000 siafunds and they can't be divided. This is for two reasons. The first is that siafunds are actually rather computationally expensive to have on the blockchain. Keeping them at 10,000 discreet units minimizes the workload. Second, siacoins are only so divisible. How do you round if you don't have discreet units? Having discreet siafund units makes rounding simpler.

But the biggest reason is that I want people holding siafunds to be invested in the health of the platform. If you own 1 or more of such a scarce asset, which has value so tightly correlated to the success of the platform, you're going to be much more motivated to help the platform grow. If you only own a tiny fraction (even if that tiny fraction is worth several thousand dollars), you are unlikely to feel heavily invested.

It also means that we don't need a giant genesis block.


1 siafund is worth the amount of revenue it generates over Sia's lifetime. If we assume that Sia is going to be dominant within 5 years, and then spend 5 years as the dominant party, and we define 'dominant' to mean 70% of all cloud storage, and we assume that in  5 years that will mean about $25b in spending on cloud storage...

1 siafund will generate somewhere around $400,000 in revenue over its lifetime. These numbers represent Sia's best-case scenario. There's a very small chance of this happen. Currently, siafunds are selling for around ~$200 each. This means, assuming everyone is using my math, people think there's a 1/2000 chance of Sia getting that big.

I think Sia's chances are a lot better than that, and I think the siafund is undervalued. One of our big risks right now is that virtually nobody knows we exist. If we got as much popularity as Storj, I think you would see the value of the siafund jump dramatically.

Thanks for the link! Shame on me for not reading last page..

..now I'm really stoked for the release on 7th, can't wait.

The answer is in post #123 on the last page.

All right, could you please describe the main differences?
I asked on Reddit already, but haven't received better answer that "it's available already".

Good observation about line-jumping and turbulence.

Siafunds are like stocks. If Sia gets bootstrapped and taking off on the market ( we have to assume that) in the future, Siafunds will be traded and measured mainly by fiats like $ by the market and the investors. Assuming Siafunds reach a market cap of one billions dollars, do you rather want 1 Siafund=$100,000 or one 1 Siafund=$1? Which one will be better for Siafunds to trade in the marketplace?

The idea comes with tradeoffs.

As divisibility increases, the implication that price can be lower increases as well (not necessarily that it will be lower, just that it can be).

It also encourages line-jumping, in the sense that if you only had .00 digits possible then there are only 100 slots in which value can trade so you must wait on a line, for what might be a decent value. If .0000, then 10,000 more slots in which value can trade, so now you must wait in line to make micro or nano-pennies - for what cause?! Will there be value added in being able to distinguish or decide if something is worth 1/10000 more than someone elses offer? Personally, I don't think so. Instead, I feel this creates nothing but turbulence, endless line-jumping, and desperation and frustration of buyers in crypto markets. This is why so many things sit at one satoshi.

It prevents many currently understood trading patterns from being followed correctly, from my observations.

When I trade a stock, I put in offer at .23 or .24 or .22, but not .22000001 or .22000000 or .22000002 because there is no value added to the system, and instead I feel it causes chaotic market actions that do not add value at this point in time. Such resolution would be worthwhile for something in which an individual unit would be worth, say, $1000000, because there is value added in such high resolution at that value. But for something worth a few dollars, it's just not the case.

I understand that these are atomic units we're trading in, but the mental application is still from dollars and cents, euro's, yuan, w/e to individual atomic units - we're just not conditioned to think on these terms though. So why create unnecessary confusion at this point in time?

Of course, at the same time, increasing the number of atomic units while keeping the decimal place in the same visual spot still serves to create the implication that there is an overabundance. So - where does the middle ground lie here between overabundance and scarcity? For now, I think 10k (now), or possibly 100k atomic units might be enough.