Technology

Belief

For years, browsers moved from being an application to a virtual machine. It is the new JVM. To achieve this task, browsers need to access more OS and hardware features. Too slowly, but still, browsers enrich their API and extend their landscape and performance. WebAssembly, webGL (and in a near future webGPU) unleash creativity based on web technology. Even if it won’t be a straight path, we believe the timing is right to build an ambitious project on such foundations.

Being Open / Interoperability

Our open philosophy will allow any next-generation dev tool or hardware to connect to our platform for the benefit of our ecosystem users. Thus Second Frontier stands as an open API platform with a standard framework supporting any current or future technology.

In practice, it means that at every stage of design we assess the best choice between openness and performance.

Architecture

Tiles

Tiles and their ownership are at the core of the SF project. Owners must have full control of tiles and be able to trade them without SF supervision, but economic proceeds must reward owners for their engagement. For that purpose, each tile is an NFT granting owners the ability to modify metadata related to its state.

In practice, each tile is an NFT minted on the Solana blockchain. It is built as a mutable pointer to an agnostic storage solution. By default SF stores tile data on an IPFS file. This hybrid solution allows maximum flexibility without compromising initial goals.

An important point is tile indexing, i.e. the ability to retrieve a specific tile from a GPS coordinate. A simple solution would be to make a proprietary algorithm to divide the Earth's surface in an arbitrary tile format. Such a way would force any application willing to participate in the SF ecosystem to request this resolution or to implement the specific algorithm, a reason why an established grid system has been a prefered solution. Geohash was the obvious choice but it led to a great distortion in tiles surface depending on their latitude. Finally H3 library has been a great tradeoff between open-source availability, ease of use and integration for SF third parties and specific qualities on GIS viewpoint. As a result, each tile has a 15 bytes id (ex: 8d393360d86093f) defining an hexagon surface of approximately 44 sqm. When a new tile is minted as an NFT, its ID is also stored in the blockchain, allowing for ulterior retrieval and independent checks.

As it would be highly inefficient to scan the blockchain to retrieve tile data, the SF platform provides an API to respond to such requests. Blockchain stays the single source of truth but for performance, SF specific data is cached and indexed in a PostgreSQL database. An important point is anyone can build such a database directly from the blockchain ledger.

Tools

Managing lands

Once acquired, land owners will have an unproductive asset and they need tools to develop it. One primary tool will be the “Stager”, it will allow users to add/remove/edit objects on a particular tile or group of tiles (called land).

Building objects

Objects will use the web-friendly and open format .gltf (or its binary form .glb). Thanks to the outstanding work provided by Khronos Group, glTF 2.0 format now brings photo-realistic rendering on the web (for instance last revision includes Physical Based Rendering, PBR), as a standard.

As most 3D softwares export to glTF natively or via addons, it is straightforward to get a particular object from various specialized marketplaces.

For some specific objects such as buildings, in the MV context, SF will provide a simple builder to place walls and facades.

In addition to these classic methods to create 3D models, the SF team is currently testing photogrammetry solutions to enable SF pioneers to enrich their land from reality. Progress in smartphone cameras and GPU computations allow to envision some applications, such as facade modeling, where limited accuracy is acceptable for a good user experience. Smartphone task would be limited to image/video capture, computation would be performed on SF infrastructure.

3D models will be stored as .gltf or .glb files on IPFS by default.

Stager

The stager will be the tool to place objects on lands. Each object will have a local position on a specific tile. Stager's main task will be to manage multi-tiles lands and check that an object only sits on tiles owned or rented by the user. Another task will be the enforcement of basic and down-to-earth rules regarding object shapes (ex: forbid construction of a 400m high structure on a 1sqm base). SF MV should stay coherent with reality standards.

Once validated, all objects and their state will be updated to the respective tiles, themselves stored on a given storage system (IPFS by default). In case of content addressable storage systems, each tile’s account on Solana will be updated to reflect the change. Implicitly only the account's owner can modify a tile.

Land manager - Dashboard

SF will provide pioneers with a web UI to drive their actions on their properties. It includes rights and fees management (rental, objects imports by third parties, …) and access policies. Usage statistics will be available.

Multiuser and social

User communication (data and voice) will be implemented with the Matrix Protocol (matrix.org), it is mature and fills many objectives: secure, decentralized and open. SF will host its own matrix node.

Each tile will have its own Matrix channel and will share its data with all 6 neighboring tiles.

At this stage, it is still unknown if Matrix’s Third Room project will be incorporated into SF, and if SF will contribute to this project, focused on MV features.

Monitoring

One particular aspect of SF is its geolocation feature. Some applications might rely on it as a fundamental aspect of their product (think a geolocalized game). Cheating could occur via location spoofing.

Thanks to a partnership with an innovative AI-based company (mainly focused on financial fraud detection), the SF framework will be able to detect fraudulent patterns in anonymized geolocation data.

Wallet

The SF goal is not to provide a wallet to its users. Nevertheless setting up a secure wallet needs a minimal amount of time to understand its principles.

SF will offer the ability for newcomers to NFTs and the crypto world to create a simple in-browser wallet to store newly acquired tiles. This wallet would store keypairs in localstorage and make sure that user has saved private key mnemonic words.

A clear disclaimer will warn that such a convenient wallet must be temporary and assets should be transferred without too much delay to a more secure mainstream wallet.

Privacy

Some major projects in the metaverse sphere are driven by companies notoriously considering users as data subjects.
Second Frontier will be designed to be GDPR compliant from scratch. GDPR is the EU regulation on personal data, maybe the strictest worldwide.

This point is particuliarly crucial as the platform will use geolocation to operate. Users need to be 100% sure that their RL moves are not tracked, otherwise it could be a major obstacle against mass adoption.

Technical team members have years of experience in data privacy and protection, especialy on the most sensitive ones: biometrics data. Partnerships with specialized third-parties are already planned to have real-time audits on Second Frontier data handling.

Solana blockchain

There are several blockchain platforms available in the market for creating smart contracts and tokens. Ethereum is still the first major player, however, Ethereum suffers from a number of problems and Solana is addressing most of them.

• Ethereum's network can be heavily congested, this is one of the main reasons why Ethereum is working for years on a new version to solve this issue. By contrast, Solana is all about speed, needing only 400 milliseconds to mine a block and this allows to process more than 50,000 transactions per second (TPS). Solana is capable of ramping up to 65,000 TPS.
• Every transaction has a fee (called "gas" on Ethereum jargon) paid to the miners processing them. Every platform applies a different fee, but it's worth mentioning that Solana is known to be very competitive. To compare apples and apples, Solana with regards to complex transactions involving NFT are much cheaper than Ethereum at an average cost per transaction of $0.00025, when it has been mostly above $30 in 2021.
  -Solana programs are designed with Rust, one of the most preferred programming languages among developers.
  -Solana is efficiently dealing with bandwidth. Using Turbine protocol, it shares the load in a manner that leads to low bandwidth use, minimal processing times, and a more resilient network that resists breakdowns.