[ANN] Sia - Decentralized Storage

[tranzactionezlive]

Can Sia censor files stored in it.  If I am a political dissident, and a country orders Sia to remove files I have stored there, what will Sia do?  

They can't - the files are encrypted, so even if you wanted to order Sia to remove the files, remove what? And how? If it's public, then they may be able to find what hosts it's on and bully them, but Sia has redundancy - they'd have to bully ALL the hosts, who are probably spread out all over the world, in order to get it to be fully removed.

Unnecessary to assume that anyone would have to bully Sia hosts to remove chunks/block uploaders by IP etc etc. Many hosts, especially those running a business based on Sia would willingly and enthusiastically remove content that has been blacklisted by a "reputable" source (read: copyright protection agency etc). Now, there is a problem and that's that not even hosts know what is stored in the encrypted chunks uploaded to them. Blacklists would have to operate on uploaders, i.e. hosts would have to block certain IPs from uploading to them.

"Sia", by the way, does not refer to an entity of any kind. There is Nebulous Labs, which develops and releases Sia code under an open source license. Thus, the question "what will Sia do?" is non-sensical. If the question was "Does Sia implement blacklisting / retroactive removal of files/peers?" then the answer is "Currently, no". I hope the Sia code-base never gets to this point. I don't see why it has to.

Sia is not under any country's jurisdiction. Still, it is entirely probable that hosts in some jurisdictions will have to comply with local laws that prohibits distribution of some kinds of content, under any circumstance eventually perhaps even when the host cannot reasonably know what they are distributing (i.e. cannot claim innocence for any reason). In those cases, the hosts themselves can readily implement their own peer black/whitelists if they so must. This would result in pseudo-private storage network within the global Sia network, where some hosts trusts only a very few other hosts, some hosts trust most except a few, and most hosts don't give a damn.

Oh, and in Edit this comment to Wolf0's answer:
A censor would not have to force ALL hosts to drop disputed content. If a censor knows about a particular file it doesn't like, then it may be in possession of a .sia file which allows the file to be downloaded from the network. That .sia file contains the IP addresses of all hosts that store chunks of the disputed file. This list is only a *subset* of all Sia hosts globally. It would be enough to target these (~20) hosts to make the file inaccessible. As these hosts may be someone's unhardened personal computer, that may not be such a resource intensive task compared to, say, bringing Amazon S3 to its knees. So, in effect, censorship by denial of service would be quite straight forward. For this reason, it would be nice if in the future an uploader can choose its own redundancy (currently hard coded) to make such an attack more difficult (at the price of also resulting in more expensive uploads).


This looks good to me.

[1714770463]

1714770463

[1714770463]

1714770463

[Taek]

Private forums are good, but don't desert this thread.

I try to check into this forum regularly, but it's a lot easier to follow the conversation on our private forum, because everything is broken up into multiple topics. I come here and see 3 threads of conversation that are all interwoven and sometimes overlapping, and it's more effort than I want to put forward to unravel it =/. And frequently, I'll completely overlook an important question that's buried.

"Sia", by the way, does not refer to an entity of any kind. There is Nebulous Labs, which develops and releases Sia code under an open source license. Thus, the question "what will Sia do?" is non-sensical. If the question was "Does Sia implement blacklisting / retroactive removal of files/peers?" then the answer is "Currently, no". I hope the Sia code-base never gets to this point. I don't see why it has to.

At this point, Sia is more of a meta-organism. Nebulous is like the queen bee (eh, kinda). We write all the code and make all the important decisions, but the community has to agree to go along with them and other people are free to write as much or as little code as they want. 'siad', which is Nebulous's maintained client, actually does support removal of contracts. But, I do think you can ask 'what will Sia do', and in asking that you are wondering which direction the decentralized ecosystem will move. At this point, Nebulous has an enormous amount of sway over the direction of the ecosystem, but that's only because you guys like us and trust us. Already we're seeing people who refuse to upgrade, we're seeing alternate clients come out for the miner, alternate graphical clients and desktop plugins being released, which means people are exclusively restricted to the software we produce.

For this reason, it would be nice if in the future an uploader can choose its own redundancy (currently hard coded) to make such an attack more difficult (at the price of also resulting in more expensive uploads).

Full programmability of the redundancy is planned. Already you can do this by forking the client and changing the hardcoded numbers. In doing so, you will not fork yourself from the network (well... not from changing the redundancy constants. There are other constants you really shouldn't change). In the future, the API will allow you to change these number per-file, rather than requiring you to fork the codebase to make the changes (one step at a time).

Unnecessary to assume that anyone would have to bully Sia hosts to remove chunks/block uploaders by IP etc etc. Many hosts, especially those running a business based on Sia would willingly and enthusiastically remove content that has been blacklisted by a "reputable" source (read: copyright protection agency etc). Now, there is a problem and that's that not even hosts know what is stored in the encrypted chunks uploaded to them. Blacklists would have to operate on uploaders, i.e. hosts would have to block certain IPs from uploading to them.

Yeah, the real protection against censorship comes from the fact that all of the data is encrypted, and nobody can tell which files are bad or which are good because every upload uses a different encryption key. You can't even make a file blacklist for Sia, only a one-time-use 'ban this guy' list. And that only works for files that are being shared. It would be easy for someone to upload a file multiple times (therefore upload it using multiple keys), and then only one copy. When it gets deleted, they upload it another time, and share another copy. It's a whac-a-mole game that won't end.

well, Sia does not provide anonymity at the moment. So a host needs only record your IP address when you upload something to be able to ban you altogether. If whatever enforcement agency decided to be iron fisted, there are probably some pretty draconian things they could do to keep illegal files off of Sia. But again, only if the uploader ever shares them. If you upload a file and never share it with anybody, it's going to be essentially impossible to accuse you of wrongdoing.

[ttx]

(not my) price forecast 0.00000089 SC/BTC
http://forum.sia.tech/topic/100/what-will-the-price-of-siacoin-be-year-end

[in-cred-u-lous]

(not my) price forecast 0.00000089 SC/BTC
http://forum.sia.tech/topic/100/what-will-the-price-of-siacoin-be-year-end

Its mine and should be taken with a (large) grain of salt. It is a bold assumption that total network storage will hit 1 Exabyte by year end. I think the potential is there, but a lot hinges on development of the protocol itself which is by far a done job.

The estimate further takes into account payment for storage alone, and not speculative elements that may drive the coin price much higher of course.

But lets just all guess for fun and don't take my estimate as any advice to invest etc etc :-)

[xDan]

I see the network is currently ~16 TB of data, but is there any graph of this over time? (to see how fast it is growing)

[Roozy]

I see the network is currently ~16 TB of data, but is there any graph of this over time? (to see how fast it is growing)

There's actually 42 TB available at the moment and 16TB being used for contracts.

I would check this out... it has a ton of graphs and useful info

http://siapulse.com/page/main

though I don't think there's a graph for storage growth over time.

[xDan]

I see the network is currently ~16 TB of data, but is there any graph of this over time? (to see how fast it is growing)

There's actually 42 TB available at the moment and 16TB being used for contracts.

I would check this out... it has a ton of graphs and useful info

http://siapulse.com/page/main

though I don't think there's a graph for storage growth over time.

Nice, thanks.

[Porsche911]

Soon more about Siacoin on INVESD.

https://invesd.org/the-next-big-trend-after-airbnb-and-uber/

Hope you guys like it Have fun

[kahir]

https://twitter.com/petertoddbtc/status/706217711999713280


^.^

[Waldozaur12]

funny name SIA

[Taek]

http://forum.sia.tech/topic/106/what-is-sia/1

[in-cred-u-lous]

funny name SIA

Apparently inspired by the name of an Egyptian god. According to Wikipedia, Sia (the god) was created from the blood spilled as Atum (another Egyptian deity) cut his own penis.

There have been some arguments about Sia's correct pronunciation. Most people seem to pronounce like "See-a", hence jokes like "See ya" on the forum while the developers insist it should be pronounced "Say-a" like in Japanese "sayōnara" ("goodbye").

So, yeah, funny I can agree with. Catchy too.

[YoyodyneSystems]

As an investment Sia looks promising to say the least.

I think it's pretty clear from the chart on Poloniex that there is a lot of accumulation going on and a very well executed pattern forming.
The curve upward towards the next rise has begun it seems.

- One of the top three current up and coming projects in it's field
- Recent mentions from top developers on the value of it's system
- Well respected and trusted team
- Interest from VCs
- 1 million dollar market cap leaves room for gains

There is no reason Sia cannot be in the league of the 10+ million dollar market cap coins.
Keep up the good work!

[americanpegasus]

As of today I have acquired a stake in Sia, and may be covering it further in the following topic: 

https://bitcointalk.org/index.php?topic=1393560.0 
 
Thank you to the devs for working so hard on a quality product.

[Forobitcoins]

Do we have a wallet that will hold siafunds yet?

I am on version 3.3 still (havent opened it for a few months)

v0.5.0 has support for siafunds in the GUI

Is there a particular process for me to upgrade?

I would hang on for a couple of days until the official 0.5.0 version is released. The available version is a Release Candidate that is rather buggy.

thanks, I'm waiting for this, I have an old client also

[bitspill]

Do we have a wallet that will hold siafunds yet?

I am on version 3.3 still (havent opened it for a few months)

v0.5.0 has support for siafunds in the GUI

Is there a particular process for me to upgrade?

I would hang on for a couple of days until the official 0.5.0 version is released. The available version is a Release Candidate that is rather buggy.

thanks, I'm waiting for this, I have an old client also

Uh... That happened already. Version 0.5.0 released a day after the post you quoted and a day later Version 0.5.1 was released.

https://github.com/NebulousLabs/Sia/releases/tag/v0.5.0-beta
https://github.com/NebulousLabs/Sia/releases/tag/v0.5.1-beta

[NUFCrichard]

Sorry for the dumb question, but how many Siacoin would someone earn by hosting?
Per GB or TB would be good to know.

I read that you need to have 95% uptime etc, but I can't find any number for returns.

[bitspill]

Sorry for the dumb question, but how many Siacoin would someone earn by hosting?
Per GB or TB would be good to know.

I read that you need to have 95% uptime etc, but I can't find any number for returns.

You can find all of the current hosts prices at http://siapul.se/page/network however lower priced hosts are more likely to be chosen and I think any host priced at or above 500 is ignored

[HYPERfuture]

Following for interest. Looks like a promising project.

[killerstorm]

Sia usage report #2.

Some time ago I tried out Sia UI, just uploaded/downloaded few mp3 files.

Now I wanted to see if I can use it for something practical: backup. So I used it to upload 3 files 28 GB total.

This time I used command-line interface as I run it on a server. I liked command-line more than GUI.

Upload speed: I don't have exact number but it took many hours to upload, perhaps 24 h total. I expected it to be faster as I'm using fairly powerful dedicated server with fast internet connection. But OK, it could be worse, recently I was a customer of an online backup provider which was MUCH slower than that.

Cost: I loaded wallet with ~60K SC, and right now it's empty. Thus at current price it cost me 60000 * 0.00000024 = 0.0144 BTC, or $6. As I understand, that's for ~1 month. In comparison Amazon S3 standard costs $0.03 per month-gb, thus 28 GB would have costed me $0.84 per month. So I'm quite disappointed, it looks like Sia is 6x more expensive than S3 standard, and 17 times more expensive than "Infrequent Access" option.

User interface: I'm quite disappointed with a lack of control over costs and redundancy, as well as a lack of status visibility. E.g. I'd like to know what degree of redundancy it has, what contracts are already paid for etc.

Stability: siad crashed twice during within 1-2 week time frame. Not cool. (I'm using this server for many years so it's unlikely to be a hardware issue.) I don't have any detailed crash report, some messages are in logs but nothing special around a time of crash. I'll try to get a better log.

Resource usage: I think siad memory usage went up to 3 GB during initial sync & upload, that's rather excessive. CPU usage is also quite high at times (e.g. during upload).

So, overall, I'm not going to actually use Sia for backups this time as it's more expensive than cloud storage providers. But it definitely looks promising.