Under the hood, Aquameta is a “datafied” web stack, built entirely in PostgreSQL. The structure of a typical web framework is represented in Aquameta as big database schema with 6 postgreSQL schemas containing ~ 60 tables, ~ 50 views and ~ 90 stored procedures. Apps developed in Aquameta are represented entirely as relational data, and all development, at an atomic level, is some form of data manipulation. Also in theory. And mostly in practice.
Aquameta is an experimental project, still in early stages of development. It is not suitable for production development and should not be used in an untrusted or mission-critical environment.
Aquameta contains seven core PostgreSQL extensions, which together make up the web stack:
- meta– Writable system catalog for PostgreSQL
- bundle– A version control system similar to git but for database rows intead of files
- filesystem– A bi-directional file system integration system
- event– Pub / sub data change events API
- endpoint– Maps PostgreSQL permissions and data acccess to the web, as well as allowing resource hosting and a web socket event server
- widget– Modular web components that can be reused
- semantics– A metadata layer on top of the database schema, for binding columns and relations to widgets, decorating keys, etc.
Aquameta is most easily installed on a Ubuntu 18 or Debian 9 instance; its installer manages dependencies by as
To install, setup a clean Ubuntu 18 or Debian 9 instance. You can run on bare metal, or using a virtual machine such as (VirtualBox) on Mac OSX and Windows, orKVMon Linux.
Once your Linux machine is setup, acquire the Aquameta package by either downloading the latestrelease, or to try the bleeding edge version, clone the repository. Then run the install script:
git clone https: // github. com / aquametalabs / aquameta.git
Aquameta has been the life project of Eric Hanson for close to 20 years off-and-on. Functional prototypes have been developed in XML, RDF and MySQL, but PostgreSQL is the first database discovered that has the functionality necessary to achieve something close to practical, and huge advances in web technology like WebRTC, ES6 modules, and more have shown some light at the end of the tunnel.
Technical goals of the project include:
- Allow complete management of the database using only INSERT, UPDATE and DELETE commands (expose the DDL as DML)
- Version control of relational data
- Remote push / pull of commits to relational VCS
- Access and manipulate the database as a file system from the command prompt
- Access and manipulate the file system as relational data from the SQL prompt
- Interal event system for pub / sub of changes to tables, columns or rows
- “Semantic decoration” allows the user to associate UI components with schema components (relations, columns, data types), auto-generate simple CRUD UIs, and progressively enhance a UI by overriding sensible defaults with custom widgets
- Pub / sub notification let peers download new content from each other as it comes available without polling
- Users communicate with each other by exchanging structured, relational data
- Decentralized P2P network with no single point of failure
- Decentralize the web
- Deprecate the file system
- Teach people to speak the language of data
- Convert word-based knowledge and information to structured knowledge and information
- Stretch a net of approximate categorization across our earth