Here is an update as to where we are, and what you can expect early next year. These are my personal opinions and expectations, as I am aware it has been a while since I personally updated you all, and there will be wider updates issued by the community staff in the new year.
Where we are
The network is now strong and responsive enough for us to start wider adoption, inside and outside of the crypto space. The beginning, non-optimal, phases of the network that always occur for newly released decentralised networks is over. I recently reviewed the launches of a number of other blockchain platforms that were written from scratch, and I personally think we scored about an 8 out of 10 — could have been a bit better, could have been an awful lot worse. For example, shortly after Ethereum launched in 2015, the recommended number of blocks to confirmation was 6000 (yes, really!), dropping to 375 after a while. They also had to perform 4 hard forks on the blockchain to fix critical issues within 3 months, as well as issuing a torrent of security updates. Bitcoin fared even worse — plagued with money printing bugs, consensus forking issues, and even acceptance of blocks without validating them. A contemporary example is EOS, whose chain repeatedly stalled , rewound transactions and went offline during the early days. By comparison, I am happy about how Genesis -> block 100,000 went for Arweave. We had some teething issues, but almost all were in the realm of software performance, not consensus/security. This is not to disparage these other projects — I personally hold each of them in the upmost respect and regard — they are truly pioneers(!), but without comparisons to other projects we would not be able to fairly assess our own performance.
The changes to the network software, and the influx of miners downloading the weave and joining the network when the main mining rewards started gave us a chance to see two of the Arweave’s core (and novel) incentive mechanisms at play in the real world. The number of miners in the network growing by 400% within a few weeks led to considerable stain on the network to distribute the data to everyone as miners clamoured to produce as many replications of the data in the network as possible (in order to maximise their personal rewards). At the time, we were laser focused on helping improve the software for the miners and making it easier to join. But with the benefit of time since this incident, I am extremely happy to be able to say this: the incentives work exactly as planned. The miners that offered faster data serving to other members of the network had more of their candidate blocks accepted, therefore they received higher rewards. Miners worked together to tune their rigs to be better Arweave miners, serving the greater needs of the Arweave network (massive data redundancy, fast serving of data on-demand), exactly as we had planned. When all was said and done, the Arweave became the data storage network that offers the highest number of redundant copies out of any available on Earth, as far as we can ascertain. This is significant because while economic incentives can be rationally argued in a clear, mathematical way, humans are innately irrational, so their behaviour is hard to predict. This unique period of massive network growth gave us a glimpse of the behaviour of the incentive mechanisms under stress, and we were very pleased with the results.
Where we are going
Now that the foundation is in order, it is time to look outward. In late January, we will be re-launching our website, with a focus on describing the advantages of, and how to use, the Arweave’s permaweb. The permaweb is the decentralised, incentivised web that lives on the Arweave. It is the first decentralised web hosting project that offers incentives to participants in the network to store the data inside it, and it is ready to use today.
Along with the new Arweave website, we will also be launching an extremely simple tool that users (inside and outside of the crypto world) can utilise to launch their own Arweave permaweb pages and applications. This tool will form the critical next step in the important process of opening up the network to new users and making it easier for them to deploy applications on the Arweave. At the same time, in order to ease adoption for those unfamiliar with the world of tokens, we will be launching a simple interface that will offer a tiny number of tokens to new users that can be used to deploy their first apps (you will be amazed how far a single AR can go). Our aim here is to become the single easiest web app deployment tool in the world, within the next few months. We will start with a friendly command line tool that will be simple for experienced web developers to use, then expand to GUI based options.
The core differentiating factors of the Arweave’s web hosting are as follows:
- Hosting is exceedingly cheap (just a few pence — fiat equivalent — per web app). Because we take a serverless approach to hosting, we cut out almost all of the overhead that makes it expensive, saving developers hundreds or thousands of dollars over just the first year.
- Zero server configuration is required, deployment is simple enough for even a monkey to do it. This factor alone saves developers countless hours of tedious and often frustrating work.
- Zero upkeep is required. Arweave web apps are forever: developers can ‘set them and forget them’. There are no recurring fees to pay, and no annoying security patches to remember to install.
The first partners are already using the Arweave’s permaweb hosting, and we will be on-boarding more in the near future. We are excited to tell you about these as they progress in the next few months.
On the technical side, we have been working recently on a new work algorithm that we eventually intend to replace the hashing that is currently used underneath Arweave’s Proof of Access algorithm. The core of this new algorithm is a random program generation and execution engine that ensures Arweave will remain a CPU and hard disk mining operation, making Arweave mining available to everyone with a simple computer, forever. There is much to say about ProgWork that we will detail in blog posts in the months to come, but the system essentially provides ASIC and GPU mining resistance by ensuring the only way to build a better piece of hardware for mining ProgWork is to build a better general purpose CPU. We achieve this by tailoring the algorithm to the strengths of existing CPUs, and focusing on things that more ‘specific’ pieces of hardware (ASICs, GPUs) find difficult.
The core of ProgWork is already implemented and has been running on a private test network since early December — stay tuned to this blog for updates.
Finally, we have also been working with a major new partner to solve their decentralised storage problems by applying the permanent storage incentives that Arweave offers to other decentralised networks that lack storage incentives. If I may quote the response of our first partner for this feature: ‘Holy sh*t, this is brilliant guys. This truly does solve a major pain point for us and many other dapps out there’. Watch this space!
Thank you all for reading! If you find these blog posts helpful or interesting, let us know and I will be sure to do more of them in the future. I wish you all a merry and relaxing holiday season.
All the best,
 There is always the possibility that secret government/corporate networks contain more redundant copies, but this information is not public. If you do have disclosable information about private systems with over 100 redundant copies, we would be delighted to hear from you — firstname.lastname@example.org.