Roadmap discussion for Vanilla 2017
It's been a year since I wrote a major update about our roadmap, and most of that discussion is outdated at this point.
There are already a few really big things in Vanilla's master branch that didn't make the 2.3 release fork last April. These include:
- Categories were rebuilt with clearer "display as" functionality and now scale really well. We added a paginated "Flat" category structure for when you need hundreds of parallel categories.
- Dashboard v3 is in, with a complete refresh of our design and functionality. It has a complete style guide and a scaffold of tools and modules for developers to more easily build settings pages for their addons.
- Under the hood, we rebuilt our plugin scaffold into a generic "Addon Manager" and have refactored the dispatcher to work more clearly and consistently. This is groundwork for a new API and self-generating API docs.
Looking ahead, we're gearing up to implement API v2, a native, internal API that will streamline addon development and enable us to run more automated testing against core. It will also allow more granular permissions, and be a more complete & pluggable external endpoint solution than anything currently available. I expect this to be complete (for some definitions of "complete") sometime during summer 2017.
In the mean time, we're also hoping to add "profile cards" support (a popup mini-profile when clicking a username), redesign our base theme (maintaining support for legacy themes), and refactor how edit history is stored (to separate it from logging).
The options I like best are to either start a 2.4 release cycle immediately (without the native API) and continue on the 2.x "point" releases, or wait for the API v2 feature and increment directly to a 3.0 release. Why 3.0? I think rebuilding all our innards & our Dashboard from scratch with a full native API counts as a product refresh. It's not a huge backwards-compatibility break nor a restart like 2.0, but I feel like it would give folks a reason to have a second look at Vanilla if it's been a while.
A year after my last update, I also worry about increasing the pace of our releases. We still have some major addons not compatible with 2.3 a month after its release and after a 6-month testing period. That leads me to believe we don't have critical mass for a more aggressive release schedule. Also, it's very stressful and time-consuming to do a major release for our small team. Hitting annually or slightly better seems like a nice compromise. The 3.0 plan gets us closer to that pace.
I'm currently leaning strongly in the direction of a 3.0 release in 9 months or so, but nothing is set in stone. Constructive feedback welcome.