Siacoin— A Leaking Chest
TL;DR. Ecology:4. Technology:5. Decentralization:3. Valuation:4. Rating:4/10
(Update 2019-02-06: Please see the semi-annual ratings review of Siacoin here.)
The Sia project was introduced on bitcointalk.org by user Taek back in 1st May 2014. By utilizing peer to peer technology, Sia wanted to create a network focused on distributed storage, where storage providers would be incentivized by block rewards in the form of Siacoin, the blockchain's native token. That was changed before mainnet launch to regular blake2b Proof of Work (PoW) with block rewards. A stand-in for the second token Siastock was issued on the Nxt blockchain under the name Sianotes. This asset represented the right to 3.9% of all storage fees paid in to the Sia network by storage renters, and were to be converted to Siastock once Sia mainnet launched. Initially around 1 200 out of all 10 000 Sianotes were sold to the public, while Taek's and co-founder nemo1618's company Nebulous Incorporated retained the rest.
In July 2014, 7.5% of Nebulous Incorporated was bought for USD 375 000 by a private investor. In October the same year, main developer Taek posted a resigned text in which he stated that he had found problems with how Sia (and other decentralized storage projects) was constructed. The team went back to the drawing board and started from scratch. On October 17th, Taek described a new, lighter protocol with enthusiasm. In November 2014, a new whitepaper was released with more details on the structure of the Sia blockchain. Siastock was renamed to Sianotes in fear of SEC taking future action against equity-like cryptocurrencies. These Sianotes were to be converted later to Siafunds with 1:1 ratio. One Siafund entitled the holder to the same 0.00039% revenue of total storage fees. But as the number of total Siafunds now were set to 25 000 instead of 10 000, holders of Siastock/Sianotes had their assets diluted. Nebulous Incorporated was to retain approximately 23 800 of all 25 000 Siafunds, meaning around 95% of of the now 9.75 percentage point 'storage fee tax' was to be sent to the company, while investors were entitled to the other 5% (or 0.5 percentage points). The community backlash was large enough for Taek to back down from the proposed changes.
By the end of 2014, a public Sia testnet was launched. Later in 2015, users could redeem Sianotes for Siafunds, the mainnet asset. On 14th of May 2015, a new BCT ANN thread was posted by Taek as the project now had a tangible product. Mainnet launched with GPU mining on 7th of June 2015. In a Reddit AMA one year later, Taek revealed that two more core developers had been added to the team, as well as two interns, totaling six persons. On 29th of June 2016, the much anticipated 1.0 version was released. Continuous releases during that year advanced the project's stability considerably. In June 2017 the first plans of a Siacoin ASIC miner was revealed - Obelisk SC1. The miner was going to be designed and distributed by Nebulous. Those plans prompted a hardfork in December 2017, with a resulting change in Sia's Difficulty Adjusting Algorithm (DAA).
As the project headed in to 2018, some controversy hit almost immediately. As Chinese mining giant Bitmain revealed their Antminer A3 model, specialized for the blake2b algorithm utilized by Siacoin, the Siacoin team threatened with a softfork. This softfork never took place, resulting in pre-sale buyers of still unreleased Obelisk miners suddenly feeling like they had been dealt a bad hand. Bitmain offered discounts to all Obelisk buyers, for reasons unknown.
In February 2018, the Sia team announced the sale of 750 Siafunds, reducing the Nebulous ownership of total Siafunds to 80%. This auction started on the 16th of April. Only 231 Siafunds out of the 750 was sold in total for a sum of around USD 1.5M. A new developer was promptly hired.
Siacoin is inflationary; currently something around 15% new Siacoins are minted to miners every year. This number will in the coming years slowly go down to 3% where it fans out. Siacoin's codebase is similar to that of Bitcoin's, and can be assumed to be a modified copy with changed mining algorithm, among other things.
The Sia network processes between 5000-10 000 transactions per day, which is not comparable with Bitcoin or Ethereum, but still is acceptable in the space at this time. Around 200 TB of storage is now rented through the Sia blockchain. The whole thing works, but the market has evidently not adapted it yet. Looking at the Sia sub-reddit, there is activity and some very good discussions about Sia's future, about the Obelisk fiasco and potential roads going forward. Overall, the community seems to be of a higher than average standard, engaging in complicated discussions.
The number of total on-chain transactions is obviously somewhat proportional to the amount of interest individuals and companies show for the decentralized storage solution that is the pinnacle of Sia development. It is still not clear that by utilizing a blockchain can total costs be reduced enough, or at all, to compete with centralized solutions provided by for example Google or Amazon. What the Sia platform can offer is a certain censorship resistance, which may attract some (shady) activity. The Sia developer team has stated multiple times that they think other companies rather than regular users will directly utilize the platform in the end. Those companies will in turn offer user-friendly storage solutions to users who don't want to store the public blockchain locally, or having to deal with other complicated processes where the blockchain is involved. Such a development could still mean problem if some jurisdiction deems some data stored to be illegal. Having companies to sue make an attack on Sia easier, even if the data stored is encrypted.
At the moment, the Sia community is rather fractured and angry for multiple reasons. Some feel that too much resources have gone in to the development of Obelisk ASIC miners, which turned out to be a much harder venture than what was initially thought. Those resources could have been used to mitigate bugs and some other crude aspects of the main Sia client instead. Since Bitmain beat Obelisk to the market, the result was higher mining difficulty and lower profits for pre-sale buyers of Obelisk miners. Adding insult to injury, the shipping deadline was not kept, and no refunds have been offered to buyers.
Lastly, no long conversations were held with regards to a potential fork that would invalidate Bitmain miners. The developer team seems to have decided on their own accord that such a fork would not be in the greater interest of the platform. A new debate about temporarily invalidating Bitmain miners have just recently flared up, and it will be interesting to see how the team responds. BD Ratings estimate the support for such a fork to be very high.
Reasons: Some on-chain activity. Active, engaging and somewhat fractured community. So far low utilization of decentralized storage.
The co-founder and developer Taek has shown honesty and a resistance to the sunk cost fallacy, as the Sia core technology was shown to be broken in an early phase of development. If such proofs surface, it is good if a project acknowledges that fact and immediately starts thinking of possible solutions. All this occurred back in October 2014 however, and although Taek back then doubted that decentralized storage solutions would be able to compete with centralized storage solutions, much has happened since and the project quickly came up with workarounds.
In an early post from 2014, founder Taek speculates about the future of Siacoin and Siastock/Sianotes/Siafund. He admits that a high inflation would mean that storing wealth in Siacoin would not be a good decision, and then argues that Siastock on the the other hand could be a very good investment since it derives value from Siacoin. BD Ratings does not understand this circular reasoning; since Siastock holders receive their profits in the form of Siacoin, they are indeed affected by the inflationary nature of that native blockchain token. nemo1618 iterates the prediction of Siacoin being a bad store of value. Users of storage can of course offer more Siacoin should they have a low value, but the premise of any value itself is greatly derived from scarcity. It would make more sense to make Siacoin scarce to incentivize suppliers of storage and miners alike.
Even if we look beyond the initial few years, Siacoin still will have a considerably higher inflation than for example Bitcoin. It is the opinion of BD Ratings that it will be hard to compete as a store-of-value protocol with a too high, never ending inflation, especially now when the cryptocurrency sphere is maturing and many older chains' inflation drop under that of gold. This economic choice of high inflation is somewhat puzzling as it doesn't fit with the team's great knowledge of mining overall. With the development of Obelisk ASIC miners, and some really good articles on the subject, the core developers have proven that they understand PoW mining in depth. This hands-on knowledge is an asset as the security of the blockchain's underlying currency partly rests upon it. The fact that the Siacoin codebase partly is based on Bitcoin's codebase is a strength as well. It means it inherits parts of Bitcoin's technical superiority.
As we lastly move on to the GitHub repository, there is considerable activity by a handful of developers. Sia recently moved development to GitLab, and anyone can inspect the activity there as well. A continuous stream of new client version have been released throughout the years.
Reasons: Active GitHub/GitLab. Real decentralized storage solution that works. Good PoW knowledge. Way too high inflation.
In a proposed move in November 2014 that came without any discussion whatsoever, the Sia team wanted to change the fee structure in a way that would have resulted in Siafund holders having their assets diluted heavily. The Sia project was IPO funded by investors who assumed a 'storage fee tax' of 3.9% were to benefit them in the long run. As the Sia team wanted to increase this 'tax' to 10% while printing more Siafund for themselves, they gravely mistreated their investors. Investors held, before the proposed change, around 12% of all Siafund, resulting in a fee tax revenue of approximately 0.5 percentage points (0.039*0.12), while Nebulous kept the other 3.4 percentage points. After the proposed fee change, Nebulous would have been entitled to around 9.5 of 10 percentage points, leaving the same old 0.5 percentage points to investors. The Sia team claimed that this was not a dilution since investors would still get the promised part of total storage fees. The overall increase of the storage fee tax to 10% would however affect the total provided storage capacity as storage owners then would only keep 90% of their profits instead of 96.1%. The whole thing feels like it came from pressure applied by the venture capital company that bought a stake of Nebulous.
Looking at some mining statistics, it is evident that F2pool.com has a worryingly high rash rate. At time of writing, it is slightly over 50%, resulting in a threat to Siacoin. It is in the mining pool's self interest to continue with the profitable mining, but even having the option to double spend is an open attack vector.
Finishing up the decentralization discussion, it should be added that the Sia team has a good understanding of what decentralization is and why it is beneficial. The Sia client for example seems to be the storage solution on the market that is the least dependent on some centralized aspects. It is also no surprise that it launched without a considerable pre-mine. There is a company in the middle of the decentralized spider web though, and it ciphers funds, albeit a small amount every year.
Reasons: Mining pool with majority hash rate. Centralized decision making that was to affect investors.
All Siacoin in circulation are worth a total of around USD 285M. That is right now a quite high number for something that has high inflation, relatively few developers, and is focusing on a niche part of blockchain technology (decentralized storage). The Sia project needs more adoption of its impressive client, further decentralization in terms of both ethos and hashpower, and a more attractive blockchain token in terms of inflation. A one-time fork to hurt Bitmain miners is probably not a bad idea as well, as it would restore loyalty to large parts of the community.
Compared to Storj and MaidSafe, Sia is higher valued.
Reasons: Siacoin is not constructed to be a good store of value (yet).