One of the most straightforward answers when it comes to the off-chain versus on-chain blockchain argument is: why even use a blockchain if you are going to store sensitive, crucial data in the traditional way? Using a blockchain to manage, authorize, and secure sensitive data is one of the key features of blockchain technology, to ignore this is foolish. As we move forward with decentralization, especially in regard to decentralized applications, keeping crucial information on-chain is more important than ever. “But the cost is too high,” claim the naysayers— the truth is they just need clearer priorities and stronger technology.
On-chain versus off-chain in brief
As the name of the argument suggests, the discussion of keeping data on-chain versus off-chain stems from blockchain companies moving user data, sensistive information, and forcing off-chain payment processing through off-blockchain means to save on costs. You’ll often hear that the reason is to increase scalability and privacy, but those issues stem from wanting to save money: save money on nodes, on mining, on development, on block validation, etc.
When blockchain companies rely on off-chain solutions they are vulnerable due to their centralization, which brings us back to the world pre-blockchain. Simply Googling “data leak” will bring up dozens of articles and news stories about centralized companies losing control over their data. Bad actors are definitely one problem, but what about losing data due to human error?
On-chain blockchain companies that store their sensitive data in a decentralized manner are not only far more protected against bad actors, they are also protected against the human error of accidental erasing, corruption, and server downtime. Dapps, in order to be truly successful, will need the benefit of all of these perks to prevail, especially dapps that deal with sensitive information: medical dapps, cloud storage dapps, financial dapps, and more.
But we already have dapps for those things, don’t we?
There are certainly options that exist that cover these intents in the dapp ecosystem, but are any of them truly on-chain at the moment? Looking at some popular blockchain solutions, projects like Filecoin, SiaCoin, and Bittorrent, we can see a trend in off-chain utilization: payment-processing and data transfer requests. According to their whitepaper, “ BitTorrent Inc. will deploy an ‘on-chain/off-chain exchange.’ The exchange will enable the transfer of tokens between a high-performance private ledger and the public TRON blockchain.” Filecoin is taking a similar approach, in that it “the retrieval market will operate entirely off-chain,” and that “miners are paid in filecoin via off-chain payment channels for quickly providing files to clients who request them.” Sia follows this trend, with “payments between renters and hosts occur[ing] off-chain, greatly increasing network efficiency and scalability.”
These off-chain methods occur for two primary reasons:
- Scalability problems - Because of the limitations of the blockchains being used (in both blocksize and block processing time), these dapps are forced to rely on off-chain solutions to manage important, sensitive parts of their business model.
- Costs saving methods - Building a dapp on an existing blockchain is cheaper than developing your own, not to mention preexisting blockchains already have nodes and miners actively participating in them.
This problem begs the question: are there any blockchains on the market that can handle everything on-chain instead of relying on off-chain solutions to round out the application?
The scalability of certain on-chain solutions
A few projects focusing entirely on on-chain solutions are worth bringing to attention in the ever-growing off-chain ecosystem.
Cardano is a blockchain that is developed by university professors and data scientists and claims to be the first blockchain platform to evolve through scientific philosophy and a research-first approach. Cardano operated entirely on-chain and does so through a few different methods. It should be noted that Cardano isn’t capable of storing large amounts of user data on-chain, instead, its purpose is as an ownership data manager, with its use-case intent being in financial operations to validate and record transactions between parties through metadata, attribution, and compliance. Although Cardano isn’t intended to store data for user retrieval, the ever-growing size of its blockchain would undoubtedly become unruly if it didn’t solve the scalability issue. To overcome this, Cardano is considering a few unique approaches:
- Pruning - Removing data overtime that isn’t crucial and needed for the every node to carry.
- Subscription - Nodes subscribe to specific subsets of data, but have to maintain a certain percentage of the network.
- Compression - Compressing chunks of data to reduce overall storage needs.
- Partitioning - Sharding the network to increase speed by reducing the network need for each node.
Overall, Cardano’s approaches to solving the scalability problems can be a lesson to other on-chain focused blockchain companies attempting to do more than just acting as a ledger for financial transactions.
EOS.io, the virtual computer blockchain, has also approached the problems of scalability in a creative way. EOS’s approach is multipronged, but some people definitely take issue with a few of the methods EOS is choosing to use. Overall, EOS breaks its scaling strategies down into three categories:
- Vertical scaling - this includes building better engines to increase smart contract performance (and reduce data waste), performing continual network surveillance to identify potential node areas for improvement, improving flexibility and throughput by increasing database bandwidth, and introducing new blockchain technologies to the EOS blockchain as they become available.
- Horizontal scaling - the main focus here is sister-chain and side-by-side based chains that rely on oracles and similar pivots to function, allowing smart contracts to execute in parallel with one another (so long as they have no crossover between them), and putting an emphasis on developing cross-chain dapps to keep key data on-chain through multiple chains.
- Data access scalability - the key takeaway of this method is to allow developers to build performant query mechanisms that will enable their complex queries to be executed server-side instead of client-side, thus reducing the overall transmission of data across the EOS blockchain.
Problems with EOS, however, are loudly voiced in the blockchain community. The most prevalent of these problems is the concept that EOS isn’t actually decentralized, but is instead a cloud-computing network controlled by a centralized network. This idea comes from the way EOS validates blocks, choosing dPoS with no Byzantine Fault Tolerance, leaves EOS subject to network hijacking through vote manipulation.
ILCoin, which has a use-case more akin to EOS but a technological underlying closer to Cardano, is different from many other blockchains for a few reasons. First, ILCoin seeks to be more than the simple technology underlying Bitcoin, without the limitations that Ethereum and similar chains suffer. ILCoin uses SHA-256 PoW for security and confirmation, but its true power lies in its RIFT Protocol. In short, RIFT Protocol creates a potentially unlimited network size by allowing mined blocks to contain mini-blocks, the latter of which are theoretically infinite in number. Mini-blocks benefit the network in two key ways:
- Mini-blocks don’t have to be mined - instead their hash is generated automatically based on the code of the main block they come from.
- Blocks reference mini-blocks, which in turn reference transactions. Because the main blockchain block size is a proven 5GB, thousands of times larger than the block size Bitcoin currently uses.
In the same way that Cardano uses local and network-wide focus for nodes, ILCoin allows blocks and mini-blocks to sync asynchronously, instead of in a queue, to speed up transaction time. This is especially important because ILCoin operates as a decentralized data storage service. ILCoin’s calls their storage service the Decentralized Cloud Blockchain, and the difference is that everything, from transferring data to making payments, is all kept on-chain when using ILCoin. They are able to do this through the aforementioned mini-block development, which can handle far larger amounts of data to encompass not only file sharing, but payments, user accounts, and more.
Given ILCoin’s lesser known status, you may want to read more about these technologies in ILCoin’s Whitepaper, Business Paper, and RIFT Protocol Blueprint documents to get a full understanding of exactly what this company is attempting to achieve.
Dapps in the on-chain future
Once we, the blockchain industry and decentralized consumers, make the shift into a completely on-chain future we’ll finally be able to trustlessly trust dapps without issues. Are you sick of reading about data leak after data leak? Because I know I am. Users should have no interest in giving companies control of their sensitive information, whether that be payment info, personal health data, personal files, precious family photos, identity documents, or anything else they don’t want to be shared with the public until we know our data is secure.
Dapps that stay entirely on-chain will increase the security for sensitive applications, and simultaneously allow users to be rewarded for their interactions in profound, micro-instance ways. Google Drive is great, but it would be even better if I didn’t have to trust Alphabet to keep my data safe. The same goes for storing photos on iCloud, connecting with friends on social media, or making payments through portals. Blockchains are changing the future, but unless we stay true to blockchain’s purpose and keep everything on-chain it may not be a future that truly benefits us after all.