[META] farmOS 2.x Roadmap

This issue will be used for the high-level farmOS 2.x roadmap, with links to child issues as they are created. Please comment on sub-issues instead of this issue whenever possible.

Problem/Motivation

farmOS is currently built on Drupal 7, which was released in 2011 and will be end-of-life in November 2021 2022. This means it will no longer be supported by the Drupal community: https://www.drupal.org/psa-2019-02-25 Updated: https://www.drupal.org/psa-2020-06-24

In the meantime, Drupal 8 (released in 2015) and Drupal 9 (released in 2020) offer significant improvements that farmOS will benefit greatly from. Drupal 8 saw a major refactoring from Drupal 7, moving away from a lot of the old "Drupalisms" and procedural code to more modern object oriented PHP standards and frameworks like Symfony. Drupal 9 is essentially just Drupal 8 with deprecated code removed, and moving forward the Drupal community intends to keep upgrades easier in the future (https://dri.es/making-drupal-upgrades-easy-forever), to avoid the same difficulties of the D7-to-D8 upgrade paths.

So, the current plan is to skip Drupal 8 and build farmOS 2.x on Drupal 9.

It is important for farmOS to upgrade to Drupal 9 to ensure that it can continue to be developed/supported into the future.

Apart from that, we also have a long list of improvement ideas for farmOS. Many of these will be easier to implement with the tools that Drupal 9 provides. And it's better to build it once in the new way, rather than rebuild it again after the upgrade.

The upgrade will have many challenges, but the sooner we complete it, the better positioned we will be as a growing community. It is better to invest in feature development on farmOS 2.x than it is to build in farmOS 1.x and then need to rebuild/refactor for 2.x. Of course, we also have to balance this with the needs of current 1.x users.

This is also an opportunity to take the lessons that we learned in 1.x and reapproach some things in 2.x. The migration path from a 1.x database to 2.x will be fully-automated, so we can make changes to the data model in 2.x and write automated migrations from the 1.x data model.

Some of the high-level goals of the upgrade include:

• Improved APIs (JSON:API / GraphQL / Auto-generated docs with OpenAPI)
• Improved translation/localization capabilities
• More granular data (entity revisions for assets, logs, areas, etc)
• Map enhancements (standalone JS library, timeline slider, etc)
• More accessible to frontend developers (Field Kit / Field Modules)
• Cleaner codebase (D9 OO)
• Improved performance (lazy loading code, more caching options, etc)
• PostgreSQL database support
• Upstream support/security (Drupal 7 EOL 2021)
• UI improvements (modern theme options, in-place editing, etc)

And some things that will most-likely be done at the same time:

• #2363393: Make "Area" into a type of Farm Asset
• Inventory management
• Comments on areas, assets, logs, etc

Proposed resolution

The following meta issues outline the major initiatives of the upgrade. Each
contains overviews and links to specific tasks:

• #3151231: [META] farmOS 2.x Dependencies
• #3151232: [META] farmOS 2.x Build Process
• #3151233: [META] farmOS 2.x Data Model
• #3151240: [META] farmOS 2.x Migration
• #3151241: [META] farmOS 2.x API
• #3171281: [META] farmOS 2.x Data sharing and access control
• #3151243: [META] farmOS 2.x Python and Javascript Libraries
• #3151244: [META] farmOS 2.x Importers
• #3151246: [META] farmOS 2.x UI/UX
• #3151250: [META] farmOS 2.x Field Kit
• #3151247: [META] farmOS 2.x Contrib Modules
• #3159886: [META] farmOS 2.x Localization
• #3162602: [META] farmOS 2.0.0-beta1