This repository is the core only. You'll also need a front-end, from the list below.
If you want to experiement, you can build just the core like so:
> cd rust > cargo build
Here are some other front-ends in various stages of development:
xi-mac, the official macOS front-end.
fuchsia/xi, a front-end in Flutter for Fuchsia, and might work on other Flutter targets.
xi_glium, an experimental GL-based front-end in Rust.
xi-gtk, a GTK+ front-end.
xi-tui, a text UI.
There are notes (I wouldn’t call it documentation at this point) on the protocol at frontend.md. If you're working on a front-end, feel free to send a PR to add it to the above list.
Here are some of the design decisions, and motivation why they should contribute to the above goals:
Separation into front-end and back-end modules. The front-end is responsible for presenting the user interface and drawing a screen full of text. The back-end (also known as “core”) holds the file buffers and is responsible for all potentially expensive editing operations.
Native UI. Cross-platform UI toolkits never look and feel quite right. The best technology for building a UI is the native framework of the platform. On Mac, that’s Cocoa.
Rust. The back-end needs to be extremely performant. In particular, it should use little more memory than the buffers being edited. That level of performance is possible in C++, but Rust offers a much more reliable, and in many ways, higher level programming platform.
A persistent rope data structure. Persistent ropes are efficient even for very large files. In addition, they present a simple interface to their clients - conceptually, they're a sequence of characters just like a string, and the client need not be aware of any internal structure.
Asynchronous operations. The editor should never, ever block and prevent the user from getting their work done. For example, autosave will spawn a thread with a snapshot of the current editor buffer (the persistent rope data structure is copy-on-write so this operation is nearly free), which can then proceed to write out to disk at its leisure, while the buffer is still fully editable.
Plug-ins over scripting. Most text editors have an associated scripting language for extending functionality. However, these languages are usually both more arcane and less powerful than “real” languages. The xi editor will communicate with plugins through pipes, letting them be written in any language, and making it easier to integrate with other systems such as version control, deeper static analyzers of code, etc.
JSON. The protocol for front-end / back-end communication, as well as between the back-end and plug-ins, is based on simple JSON messages. I considered binary formats, but the actual improvement in performance would be completely in the noise. Using JSON considerably lowers friction for developing plug-ins, as it’s available out of the box for most modern languages, and there are plenty of the libraries available for the other ones.
This is still a project in its early stages. The Mac build has basic editing functionality (it was used to write this README), but looks very spare and is still missing essentials such as syntax highlighting and auto-indent. At the moment, it’s expected that its main community will be developers interested in hacking on a text editor.
The main author is Raph Levien.
We gladly accept contributions via GitHub pull requests, as long as the author has signed the Google Contributor License. Please see CONTRIBUTING.md for more details.
This is not an official Google product (experimental or otherwise), it is just code that happens to be owned by Google.