Topic: Sia - Siafund Redemption Deadline: June 1st, 2015 - page 24. (Read 68743 times)

Interested.

Interested in this Innovation project and idea.

This looks promising  but I suppose there will be quite a few bummers on the road.

I hope the devs are bunch of strong willed f*#kers

interested

We'll be notifying everyone who's expressed interest at least 2 weeks before we do any sort of IPO.

Is it possible to get notified when the IPO starts? With a pling in my bitcointalk-inbox or likewise ...
Interested as I am, it won't be possible to check this thread every day.

Thanks for stopping me from putting money into StorageCoin I was just about to...

And, hey, let's stop talking about child porn... it is irrelevant to crypto coins. It's like connecting the invention of electricity and people watching too much TV, they are unrelated. The dominant coins of the future will be anonymous and BitCoin will be put in nursery.

Sorry about that. The IRC channel has been inactive for a while because for the most part it's just 2 people living in the same timezone. After going a few weeks with no messages, we stopped checking. If this starts gaining momentum however we'll bring it back.


Looking at the maidsafe results, what we might end up doing is selling 'tokens' for 0.001btc each. We'd sell for a week or so, and then at the end we'd divide X% among all the tokens that got sold. Or maybe we'd set up a dynamic price, so each token is worth 0.0Y%, but the cost of a token changes as people buy them up or do not buy them up. We'd want something that would end up being fair, such that everybody interested got siastock and had at least 2 weeks to get in. I'm really not sure what the best approach would be. For those curious, we'd be wanting on the order of 200 bitcoins. We'd be happy to accept more of course, but less wouldn't give us enough money to work on Sia full time. If somehow things exploded and we ended up with 500 or 2000 bitcoins, we'd probably hire a few extra developers to help us out in the spirit of trying to release quicker. We would also put some money aside for security audits of the beta.


(better to ask that question now than to wait until it happens)

That's a good question. We are legally required to do everything in our power to take CP and other illegal files and DMCA'd content off of the network. But just like the creators of bittorrent, Sia is a protocol so I don't know how responsible we'd be for what happened. In the worst case, the NSA forces us to back-door the protocol such that we can censor any file. I don't think this is a realistic outcome.

The US Government has been playing nice with bitcoin. Coinbase, Bitstamp, Bitcoin Foundation, etc. have not so far been held responsible for the way that people use bitcoin. I'm under the belief that Sia would work the same way. If child porn ended up on Sia, the burden falls upon the uploader and the people linking to the uploaded file. For many reasons, I think it would be very bad to enable built-in censorship of files on Sia. If we were pressured to, we might even be able to get some human rights groups to fight for us in court. I have no idea.

Another example is Tor, a technology that has enable child porn in particular for a long time. Tor is a legally protected application. It's not the Tor devs or the relay nodes that are child pornographers, and they don't host Tor to enable child pornography. People who upload illegal content to Tor are not protected, but the ecosystem is protected, and they do have powerful legal groups ready to fight for them.

Censorship is always a difficult thing to talk about, because nobody wants to enable activity like child pornography, least of all us. At the same time, we don't want to enable the government to censor things like criticisms of the people in power. As far as Sia is concerned, if you enable one you enable the other, and you are left trusting that the person in charge of censoring child porn will choose not to censor other actions like criticism of the government.

We very much want to be viewed as "good guys" in the eyes of our government (which happens to be the US government). I have no idea what this will entail or what court decisions might be made. I do think that Maidsafe, Cryptosphere, Datacoin, and other such projects are going to run into the same exact problems.

If I recall, wasn't child porn stored in the Bitcoin blockchain? Didn't someone find some way to encode an illicit photo into a set of transactions? I know other things (like Bitcoin's whitepaper) have been encoded into the blockchain. What is the legal response in Bitcoin's case?

I read your whitepaper a while ago and was intrigued by the ideas presented. Seems like storJ, maidsafe et al have a little competition. I did join a while ago to the IRC channel but there was no activity there at the time.

I hope you learn from the lessons of the encumbered maidsafe IPO for issuance of Siastock. I think a decentralised fundraising on a platform such as CounterParty, where users can directly send BTC, and later trade shares with no requirement for proprietary MSC style tokens  would be a good fit.

We're expecting throughput for a file to be on the order of gigabits per second. Latency on low-redundancy files will be on the order of 100 milliseconds. So, you could theoretically boot your operating system from Sia instead of a local hard drive, given that gigabits per second is close to what SSDs can achieve. The 100 milliseconds isn't typically a problem, but could be for disk-intensive applications.

So the biggest bottleneck is probably the price of bandwidth. If you're a corporation like Google, you probably wouldn't want to use Sia to process search requests, as I'd imagine your bandwidth prices would be too high.

But anything that's already going over a wire would be fair game. Imagine if all images on the web were embedded directly from Sia. This could reduce hosting costs for many companies without degrading the user experience. Especially for video websites, this would be a huge deal. Right now hosting and streaming video is very expensive, and Sia directly addresses these costs without adding any additional expenses with regards to bandwidth. Pretty much any static webpage would also be supported.

Backups like Dropbox are an obvious use case. But applications could also talk directly to Sia. Save-data from video games could use Sia, your music library could stream directly from Sia, and your movie library too if you have a strong enough internet connection.

Interested in this great project and idea.

Interesting, I really like this idea. Useful work instead of performing wasted computations is definitely going in the right direction.

What are some use-cases for this? For example, if I owned a company, I would still want internal data storage within my company, if anything, to save costs on bandwidth. The only applications I can think of (right now) for this kind of distributed storage is similar to what I would upload to dropbox, which is free up to a certain point. And, I suppose, for those concerned about censorship.

Pandaisftw

I will be monitoring this thread with my both eyes.

Sia is the Egyptian personification of wisdom, or something like that. It ended up not being exactly what I thought it was when I originally picked the name, but I like it nonetheless. It's not an acronym or anything, just a single syllable.

What is Sia? Security Industry Association

Interested.

"network will mine coins such that the annual inflation rate is kept permanently at 5%"

No maximum. Every year the coin volume will go up by 5%. We do it this way because we want the coins to be about the storage, not about the speculation. There will be explosive growth in the first 4 years however, so there's still a ton of opportunity to jump on the coin early.



It's a lot like bittorrent, but instead of each file being entirely stored on each seeder, each file is only partially redundant (you get to pick the redundancy). It's hard to guess at exactly how much network traffic there will be, but probably a lot for people who are hosts.  There will be support for capping your bandwidth, though this will affect your income. (The model for charging for bandwidth isn't worked out yet).

Any file can get into storage. The clients will encrypt them first, but the network doesn't care. If you want to upload an unencrypted file (so it's available to the public), you will be able to do so. I'm not sure what you mean by "crappy" file, but it's just storage. If someone uploads a virus, it can't hurt your computer because the client doesn't open or read files, only stores them. If you try opening it with some program however, it might infect you.


1. Maidsafe has been working on this since before Bitcoin. A lot of solutions that are now available and have been thought about in greater depth were not available when maidsafe started. I'm pretty sure that if they started now, they would get a lot farther in the same 7 years.
2. The maidsafe project is considerably more ambitious than Sia. They are doing proof-of-bandwidth, proof-of-storage, proof-of-computation, and something related to uptime monitoring. It also seems like they intend files to be mobile (?). Regardless, from the best that I can tell, Maidsafe is much more ambitious than we are, and takes a more complex approach to many of the problems. This will increase the development time.
3. This project is still very ambitious, and is a few months away from a prototype. We are definitely trying to do some very difficult things. We are not asking for money at this point, nor are we giving away coins at this point. We plan to do both in the future, but will wait until we have something more substantial built.

I know that a lot of you are cautious about scams. If you look, you'll see that nowhere in this post or on our github page does a way to pay us appear (no bitcoin addresses, etc.). This project is very real and is very much not a scam; we aren't asking for your money.

Do we have a proof of concept yet?:

Well, we have a working demo that does not a whole lot. This currency is based on quorums, and we just finished building the base for quorums. A bunch of hosts can get online, engage in communication (following Leslie Lamport's solution to the byzantine generals problem) and kick dishonest hosts off the network. At the moment, there is no support for joining with a wallet or storing files. That's what we're working on now. Stay tuned.


Each set of 8GB that a miner offers to the network becomes a 'participant'. Each participant downloads and watches a portion of the network, and is paired with 128 other participants (chosen randomly). The 128 operate together reaching consensus using the Byzantine Generals solution provided by Leslie Lamport in his paper "The Byzantine Generals Problem." The pdf can be be found in the doc/ folder. We used the signed solution, meaning every honest host in a quorum is guaranteed to reach the same block as every other honest host within the quorum.

Quorums are arranged into a tree. Each quorum has a set of sibling quorums that all share the same parent.

When a block is found, it's announced to the sibling quorums, and to the parent quorum. The exact algorithm is not pinned down, but each participant will send the block to 2 or 3 other random participants. This keeps communication to a minimum. A group of dishonest hosts could announce a dishonest block, so there is protect against this; upon receiving a new block, each host will fire off a confirmation to the other participants in the announcing quorum. With very high probability, at least one honest host will see the dishonest block, and will be able to alert the receiving quorum about the dishonesty. After we pick the exact numbers, we will write a formal algorithms with proofs demonstrating what attacks are possible and with what probabilities the attacks will succeed. (Our target is to make the attacks as unlikely to succeed as double-spends on the Bitcoin network when an equivalent volume of hosts are dishonest).

Does that make sense? It's very different from any existing model, using a bunch of communicating quorums instead of using a single large blockchain. The primary reason for doing this is because Sia will have a lot more information in the blockchain than Bitcoin. For example, the header data of every single file on the network is stored in the blockchain. If every miner needed all of that data, you'd be looking at hundreds of gigabytes even before figuring out which miners had which files. A tree was absolutely needed.

We believe the algorithm to secure, but that needs to be confirmed by others. At this point, the algorithm is not formally written down, which makes it hard. Expect to have something more concrete regarding quorum consensus in the next 2 weeks.

HERE let me throw BTC at you.

In this system, I wonder how the blockchain consensus will be reached.

I always support the cautious approach. But did you check the github, the sia-dev group history, and the general profile of the devs (who don't appear to be anonymous)? Also bear in mind the summary does state the release is a long way off. Agree that the problem sia is looking at is not an easy one, but if it's an outright scam it's a very elegantly engineered one!

That sounds really great. But  it isn't appropriate   to advertise here .