We primarily offer documentation, peace of mind and expertise to our users - and code/configuration is simply an implementation detail. Well structured documentation aimed at different audiences is essential to 2i2c’s long term health. This document lists the audiencs we serve, what kinda docs they might expect, and where we provide them.
For each feature we add or change we make, we should create & update documentation for all of these audiences, as applicable. There are a few key audiences that are outlined below.
The general public¶
People visit our website to learn about us, and investigate if we could be of use to them. Communicating what value we can add to them is extremely important, so any feature we write should be integrated into our website. The focus should be on the value we can add to potential users, and should link to other kinda documentation for more detail if needed.
Ultimately, our hubs are built to serve researchers, students and instructors. These are our end users, and they need to be able to get their work done.
The infrastructure we provide is primarily opinionated bundles of upstream software, so we don’t need to rewrite documentation for all the software they might use. However, since we make opinionated choices about JupyterHub deployments, some repetition of upstream documentation that assumes the specific choices we already make is helpful.
We should provide at least the following kinds of documentation:
Tutorials for common workflows on our infrastructure, since this is heavily driven by the opinionated choices we have made while building it.
How-to guides to help users accomplish a specific well defined task, especially if it is something they know how to do in a different system. The documentation titles should always be of the form How do I
Component reference to inform users where to find documentation for the component they might be having issues with. Most users are unfamiliar with the intricate details of what component does what, so might not be able to find the appropriate place to look at.
Hub administrators are the interface between the 2i2c engineers running the hub and the actual users of the hub. They are usually a part of the community using the hub, and are interested in the configuration of the hub infrastructure as well. They should be aware of possible configuration options they can choose to serve their community best, and be empowered to make those choices without interaction with 2i2c engineers wherever possible.
They help make decisions about the configuration of the hub that benefits its users most, and hence we should provide at least the following kinds of documentaiton.
Configuration guides to help them pick the configuration that will be best fit for each of our major features - such as authentication, kind of user interface, etc. Their hubs are then configured with these choices in collaboration with 2i2c staff. Each guide should mention how this implementation can occur - possibly with a link to documentation for 2i2c engineers .
How-to guides to help admins accomplish very specific tasks during the course of hub usage. Their titles should always be of the form How do I
These are folks tasked with building, maintaining, debugging and fixing 2i2c infrastructure. Documentation targetting them should be in 2i2c-org/pilot-hubs. repository (2i2c-org/pilot-hubs), regardless of the kind of hub it relates to. Since we run our hubs openly, with best practices that can be adopted by anyone, we should try write these to be as accessible to non 2i2c staff as possible - no secret sauce, minimal 2i2c specific process.
Here are some contexts where they would need documentation.
Tutorials, to help orient folks who might be working on areas they aren’t already familiar with. This should have clear links to pre-requisite knowledge and how it can be acquired.
How-to guides for performing common tasks that have not been automated yet. Their titles should always be of the form How do I
References, describing in detail parts of our infrastructure and its configuration. This should ideally be kept in the code describing our infrastructure - for example, terraform-docs for terraform code, JSON Schema based document generator for YAML, etc. This helps them be in sync with what we are actually doing.