Topic: [ANN] Sia - Decentralized Storage - page 363. (Read 1382191 times)

What response (if any) do the sia devs have to this claim? Follow the provably scams link for his analysis.


He seems to be saying (among many other criticisms) that the economics of decentralised cloud storage systems can't work, and storage providers will end up using dedicated servers and smaller opertions will get shut down by their ISP's. He's a wacky guy, but not a light weight thinker. Does he have a point?

I'm new to SIA and getting an idea of how the project is going.

Is there a good vibe in the camp at the moment? How is the project coming along?

Thanks

Another feature users might like is to be able to set the Hosting Price in line with the competitive average.
I see the 'Estimated competitive price' continuously fluctuating (I assume based on the network available storage and user rates), it would be good to set a +/- variance against this open market rate, so if Im new and want to pick up contracts quickly I can set 'recommended market -5%', or if Im established set 'recommended market', or if perhaps I want to shrink my storage (without dumping all my contracts) set 'recommended market +5%'.  I suppose in the early days like we are now, you need an absolute minimum, or everyone sets a percentage and the bottom falls out, perhaps the minimum hosting price should still remain.

This also saves me missing out on lucrative contracts if market conditions means space becomes short (Im not undercharging), and ensures I can pick up contracts if there is spare capacity (Im not overcharging).

oh. Technically, you can edit the duration, but if you pick a duration that's too long, hosts will reject all of your file contracts. I think the hosts are set to max out at 8 weeks by default. It might be 12, but especially while the number of total hosts is low, it's probably better not to play with that number.



A non-authorized actor can operate the core functions of the wallet only if malware has access to the machine. For the most part, it's not a concern. The feature was added for high-priority targets like Poloniex, and wasn't really aimed at everyday users. The host, especially further in the future, is going to need to be able to spend funds out of the wallet, and therefore must have access to the wallet's keys. (the host will need to put up collateral to get file contracts - collateral is important to the security model).

--------

New release candidates are ready. This means that they might be buggy, so use with caution. It is believed that all components work, including the miner.

I thought there was no debate, leaving the hosting wallet permanently unlocked is a risk to the extent a non-authorized actor  can operate core functios of the wallet.  Whether I keep 0.0000001 or 1,000,000 SC in there, there is a risk.  Sorry if my chioce of words was poor, no offence intended, but if Im going to use Sia to store my files, host for others I dont want to be shuttling funds away/back to a cold wallet all the time.

In that case you should update your v0.5.0 API documentation. In the section /renter/upload/{siapath} [POST] it states:


Source: https://github.com/NebulousLabs/Sia/blob/master/doc/API.md

My interpretation was that 'duration' therefore is customizable to any number of blocks.

What security hole are you talking about? We do recommend keeping most of your coins in a cold wallet (unplugged from the internet, and backed up in several places), and only keeping coins you need in the near future on an open machine.

v0.5.0 is aiming to make Sia a substantial improvement to Amazon Glacier. We're trying to work out all the kinks, and I think we've come a long way since v0.4.8.

You can drag-and-drop the 'siac' or 'siad' binary into a command prompt. The v0.4.8 binaries can be downloaded here: https://github.com/NebulousLabs/Sia/releases/tag/v0.4.8-beta


That is incorrect. The big change is that files will automatically re-upload, meaning that they'll stay online so long as you continue running the daemon. For v0.5.0, files will go up for 4 weeks by default and renew when there are 2 weeks left. The numbers are small so we can see how the hosts + renters behave on a short timescale in a live environment before switching to bigger file contracts. For v1.0, I think the upload time will be 6 months (26 weeks), renewing when there are 12 weeks remaining. We will probably add customizability to those numbers, but those customizations are not ready for v0.5.0.

Thanks for the replies, Im seeing this is an active community which is key for success.  I also would like to use Sia as archive, I use Amazon Glacier for that now but I think Sia could do that better/cheaper.
Sorry to be a fool, where is/how do I launch Sia as command line for Windows?
You gotta plug that security hole with the unlocked wallet, for now Ill keep my SC on one machine always locked and set the other as the one available for hosting.

Right, currently you need a backup of both your wallet and the /renter directory (where all the .sia files are).


In v0.5.0 coming soon you will be able to set your own period. So, you could increase the default 6,000 blocks to something much higher.

Running the client regularly, however, ensures the redundancy of your files on the network. As hosts come and go, redundancy naturally drops and the client is set to automatically repair files. If you don't run the client you loose redundancy and in the worst case your files become irretrievable.


There is a Gnome shell extension that sort of does this. Eventually, I would expect other OSs to follow suit. What you are asking for is all possible with the Sia API.


As the number of participating hosts grow the need for redundancy falls. Basically, down to x1.5 with about 50 hosts I think.



ShapeShift does not currently support Siacoin. I have asked them N times to consider it, but I guess they receive hundreds of requests like that per day.


Absolutely! I'm sure there are 3rd parties working on this as we speak (type).

All great suggestions me think!!

Here are a few ideas to make Sia more attractive for renters and get more contracts filled.

It's been mentioned a few times, but persistence of files across clients is a real deal maker/breaker. This is in line with what people expect of cloud storage. The holy grail is storing files in a HD brain wallet. That would turn Sia into the most powerful cloud storage system ever, full stop. It will be an instant hit among crypto enthusiasts, journalists, etc...

Also, If I suffer a hard drive crash and lose my wallet files, I shouldn't also lose access to all my stored files, in fact, I kinda used Sia to prevent such a thing from happening in the first place. Having to back up a wallet file in addition to saving my files on Sia is counter intuitive. People will get confused, lose files and freak out if this is not sorted out.

Also on the front of avoiding catastrophic failures, extending the renewal system. If you could preload a given file to be saved for say 1 year and then don't have to worry about keeping the daemon open, that would be fantastic. If I want to use Sia as a long term archival database (and I actually do), I sort of want to stash my files and forget about it without having to remember to run daemons and port things over when I change computers or have a friend use the app if I go on an extended holiday. At the very least, an independent warning system (email?) should be put in place that reminds people of the cutoff date, though I'm not sure how you would do that. The varying price is a challenge to ensuring long term reliable contracts, but there is a need for long term hands free storage even just for the sake of convenience. Maybe an intermediary dapp/contract that you can over provision with siacoin and whose task it is to keep your contracts going as long as there are funds?

Using Sia as a folder in your OS, in a way that hooks into the file system in a proper low level way like a symbolic link or a remote drive or something. Also in line with expectations around this kind of service, and just necessary and mega useful for all kinds of applications.

Not having to upload the file once for every redundancy factor unit. If there is absolutely no way around this, perhaps a bridge system could be devised, where the files are transferred to a carrier service that is then charged to distribute it to the appropriate redundancy in exchange for a Sia fee. The difference between uploading 100mb and 500mb is already too much for many/most people.

It's true that splitting the renter and host apps is a good idea. Segment and target your markets in a more focused way. Kind of like how it makes sense to split Bitcoin wallet and miner apps the same way those communities can overlap but often differ.

In-client Shapeshift.io integration. That would make the wallet more coin-agnostic and friendlier for people with Bitcoin that might not be familiar with alts.

In the same vein, a commercial service running on Sia that lets people pay in btc or fiat. Such an offering could be an instant hit depending on how attractive you can make it. The prices could be adjusted in real time according to the offer, guaranteeing the absolute best deals. That would really help drive demand.

v0.5.0 has support for siafunds in the GUI

I almost think that we should have a separate client entirely for hosting, with a separate wallet that feeds income into the primary wallet, that way the host can be unlocked all the time without putting many funds at risk. We'll have to work through that more in the future though.

The API actually supports users choosing their own password, and if you use the command line you can achieve it. I want to protect users who are likely to pick bad passwords, so I want the task of choosing your own to need to be deliberate and inconvenient. The command for choosing your own password is 'siac wallet init -p {password}', but you can only choose your own password if you haven't already initialized the wallet.

What do we have to do to support NAS? At this point in time, we have a much heavier supply side than we do a demand side, so we are very heavily focused on increasing demand. We don't have a problem with storage volumes at the moment, even though lots of people have been having trouble with ports, incompatibility, etc. I know it sounds weird, but we don't need more storage unless people are uploading more of their files to the network. We will of course continue to improve the hosting experience but the renting experience is much more of a priority.

I think the next version allows Sia to be minimized, but I'm not positive.



No paper wallet for Sia.

is there a paperwallet for sia?

So lets file it until a time when there is less heavy lifting for the devs, but I could imagine giving users even a perception of control over what they are storing would be popular.

Do we have a wallet that will hold siafunds yet?

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

Sia, the only coin without a mining pool.

Right :-)

Sia: Still no mining pool ^TM

I think its an interesting proposition: be able to set a discount for organizations your support. It would require some changes to the protocol for sure. The simplest 'fix' would be to change the hosting charge on the fly as preferred IPs connect. But this would mess up the host selection algorithm; when uploading hosts are heavily biased by cost, so cheaper hosts are preferred before more expensive hosts.

I guess that as we see Sia being developed, there will be more options to configure and use Sia per-user-preferences. Currently, we're locked into a bunch of defaults, but I'm sure adjustable parameters will be opened up for customization gradually in the future. Whether we see this is up to the devs, but I like the idea.

still no mining pool for siacoin?

Im not suggesting that I be guaranteed to store a copy of the data for the user I set as preferred, only that if I do the contract will be set at a discount.  Imagine a cancer charity signs up for Sia, 25% of the available storage hosts may be prepared to get a reduced return if they host any of the data.  Im assuming the charity would have to make 6 contracts per set of data (one with each user storing the data) and would therefore have a 25% chance of getting a lower rate than on the open Sia market.
If Im way off track, no worries, just a suggestion.