Sia - Siafund Redemption Deadline: June 1st, 2015

[Taek]

Wow! I guess it's been a full week at this point.

Things have moved along well. We're pretty close to an alpha, and pretty close to working on files.

What you can do:

1. bootstrap a quorum and add siblings using wallet scripts
2. The default bootstrap wallet premines 15,000 coins and has a script body that allows arbitrary people to request new wallets that have 100 coins each. This can obviously be abused, but for the alpha we're not worried about it.
3. The default script body for wallets can send transactions and join the quorum as a sibling
4. Wallet weights work, and you can allocate storage (kinda, there's no script that allows you to do this, but that's pretty trivial. The real problem is that you can't upload data yet).

So, apart from a few things, we have pretty much a full cryptocurrency working at a single-quorum level, which is all we're looking for at the alpha level.

What we still need:

signatures - nothing uses signatures/public keys yet except for siblings, which means that anybody can perform actions upon wallets. This should be done in a few days (Luke's working on it)
nonces - scripts submitted once can be submitted again. This is a huge security problem! This will be fixed in a few days (I'm working on it)
uploading files - to upload files securely, you have to go through a staged process. Right now I'm working on the staged process (this is actually the same code that enables nonces). After that we'll need to implement the remaining erasure coding (erasure coding is already implemented at a broad level, so the new erasure coding should be easy). This should be done in about a week.
rewarding participants - this should be super trivial
having a max volume of data stored on quorum - also pretty trivial
proof of storage - most of the framework is already in place, should only take a day or two once file uploading has been finished
non-expert client - right now you pretty much need to read the code to figure out how the client and server work, and beyond that you still won't understand most of the output. By the alpha we'll have a client and server that explain themselves and come with a guide. But it will still be pretty far from intuitive or friendly.
better testing - everything is pretty buggy. The codebase is good but we've lightened up a lot on the testing these past 2 weeks. This means a lot more simple errors have slipped through. The initial alpha release will probably not be much better, and will have occasional bugs that may crash the program and kick you from the quorum. Pretty much all of these bugs should have a 1 line fix.

So, we're getting close to an alpha. Things are looking really good. You can test out what we do have by forking us on github or just cloning the source repo. There's not much to play with at this point but the scripting is really far along. If you're feeling adventurous you can try writing your own script bodies and script inputs. The alpha will hopefully have about a half-dozen examples.

[PondSea]

Thanks for the update

[ownerbest]

Thanks for the update.
Sia is a big thing, devs are very responsible guys, I never doubt that.

[Taek]

So, I've got good news and bad news.

The good news is, I've realized that we can use the idea of a merkle tree to make the maximum sector size essentially unlimited, which saves a lot of gymnastics for larger files, and makes file proofs a lot cheaper in terms of bandwidth.  It also makes changing parts of a sector reasonably priced without needing to download the whole sector. (before, you needed to download the entire sector before uploading any changes. Now, you just need to download log(n) hashes).

The bad news is that I've got to redesign and implement a few things to make it possible. But it makes file proofs more bandwidth efficient by a factor of 200, which is a pretty big deal because that was a bottleneck before and makes the minimum requirement for being a participant much, much less. I'm still hopeful about July 1st.

[voldemort628]

But it makes file proofs more bandwidth efficient by a factor of 200, which is a pretty big deal because that was a bottleneck before and makes the minimum requirement for being a participant much, much less. I'm still hopeful about July 1st.

doesnt sound that bad after all

[grandpa_seth]

I like it when the people I've invested in tell me the bad news is some extra work to implement the really good news. I'm going to buy a happy meal now.

[robinson5]

Hi I'm very interested in Sia and wondering how to get involved.  If all the Siastock have been sold then what are the Sianotes being sold on NXT's AE?  Also, is Siacoin available yet to be bought or mined? How many Siastock have been sold?

Thanks for explaining!

[Taek]

Not all of the Siastock has been sold. If you look at this page: http://www.siacoin.com/sianotes.html, you will see the following statement:

"One Sianote can be directly converted to 0.00039% of host income from renting storage, or 10,000 for 3.9% of all host income from storage. Currently, there are approximately 1250 sianotes in circulation, or just under 0.5%."

For the near future, no more Sianotes will be released into circulation. While there are 10,000 Sianotes total, there is no guarantee that the final host income fee will be 3.9%. We are committed to having a fee of at least 0.5%, because this is a fee that we have promised to our investors. We will probably have a fee that is higher, because then we have some income for the company (which will give us power as a corporate entity to hire developers, market our product, purchase security audits, and potentially many other things that require a strong financial base).

We made 10,000 Sianotes to give ourselves flexibility to release more in the future if we decided without needing to create a new asset on the AE. We have no commitments to put more Sianotes in circulation than are already in circulation.

I think this covers everything you asked.

[Taek]

http://www.reddit.com/r/Bitcoin/comments/28uxkg/bitcoin_crowdfunder_swarm_launches_raises_720000/

Wow. Makes my crowd fund feel pretty weak. I hadn't even heard of these guys. Reading the Reddit comments though, it seems like there's a lot of skepticism around the fundraising, and people aren't sure if it actually happened, or if they spoofed the first volumes.

[Taek]

Status Update

We're nearing completion of the alpha version of the server. Right now Sia as a whole is around 9kloc, and most everything that's left to implement in the server is proof of storage. The server now supports file uploads, but the client does not, which means you still can't upload files. I'm optimistic about having the client portion of uploading done today, which means that Sia will be mostly complete as an alpha. After that's done we'll need to do a large code review, then we'll need to revise the whitepaper, and then we'll need to create a bunch of documentation around using the alpha.

It still seems like all of this can be finished by July 1st! Because it's an alpha, I'm expecting it to be a bit buggy, and it's definitely missing a lot of mechanics that will be in the full version. But it should give a reasonable idea of what the full version might be like.

The alpha is only going to have a cli server and client.

[gvans]

Thanks again for updates. The alpha will be open for anyone?

[Taek]

Yes, the alpha will be publicly available.

[Momimaus]

Can´t wait for it. 

[Lordoftherigs]

This project have mighty Devs

[PondSea]

Wow crazy James from nxt has over 1/3 of the ipo shares 

[gs02xzz]

Wow crazy James from nxt has over 1/3 of the ipo shares  

Actually James bought 1/3 of the pre IPO notes and there is only about 12% of total notes was issued for the pre IPO. Overall, James asset own about 4% of total Sianotes/Siastocks. That is still huge. If Sia is taking off one day, James asset owners will be rich.

[loveyouforever]

Will it have further IPO when it launches?

[MingZhou]

Will it have further IPO when it launches?

Based on what dev said, maybe and maybe not.

[PondSea]

Hi Taek，

Is there any indication when the alpha might be out?

[Taek]

Hi, we're a bit behind, but today we're working nonstop. It should be out before we go to sleep. We've got most things ready.

Thanks for the patience, sorry about the delay.