[last updated : 28 Jan 2018]
What Are We Waiting For?
↳ We're waiting for a few things. Without these things, the next FtB won't happen.
↳ Don't want to wait? Host FtB2 on your own server then!
↳ Want to try out what's currently available? Use FtB575!
"mu" database revolution
What is it?
↳ An eternal, backward-compatible and forward-compatible database system.
What is it? (lengthy)
↳ mu is an amorphous database service. Each database entity is self-contained. An incredibly loose service is required in order to interact with mu, since the database has no data-related constraints to enforce. To make such a system practically useful, some basic amount of conveniences are built into each entity and are managed by the server. Multiple versions of FtB can work alongside each other and share data, even without being explicitly designed with this in mind.
What's so difficult about it?
↳ Accommodating all past and future database needs.
↳ Uses self-contained entities, which seems to be the sole solution.
↳ It involves writing an unconventional database from scratch.
↳ Done, but will need to be tuned as it undergoes various stress levels.
↳ Because the database behaves equally for everyone (no official staff members), a moderately extensive permissioning system must be intergrated into it.
↳ Uncertain whether the currently implemented features are simple enough and comprehensive enough.
How's it going?
↳ mu is completed, but is currently being tested in order to ascertain its usefulness wth regard to the built-in, server-managed features. Currently, it seems likely that a simplified model will need to be designed.
"moonlight" circuit-based modular interface
What is it?
↳ A modular interface structure that bases its behavior off of circuitry.
Why is it?
↳ Strictly defines and exposes all data pathways.
↳ Encourages interface code reuse.
↳ Will allow technically-savvy users to extend existing interfaces in a standardized way.
What's so difficult about it?
↳ Figuring out what it even looks like, if it's even possible.
↳ Modeled after circuitry.
↳ Making the structure simple enough while still being performant.
↳ The fourth version finally got it right.
↳ Making a structure that isn't an unflexible mess (a.k.a. KERN3).
↳ Modeling it after circuitry to an extreme degree made it easy to reason about.
How's it going?
↳ The designs are complete. An implementation has been deployed since late December 2017 without any irreconcilable issues.
externally hosted potentially copyrighted files
What is it?
↳ Figuring out the repercussions of making charters upload their audio files elsewhere and then providing links when submitting charts.
Why is it?
↳ In order to reduce legal responsibility - not being a file hosting service.
What's so difficult about it?
↳ Dealing with file buffering, which we have relatively little control of, currently.
↳ Earnest research not yet started.
↳ Figuring out which external site would be reliable enough to recommend to users. Dropbox has already been shown to rate limit, as shown through a chart submitted by Thing O Doom.
↳ Earnest research not yet started.
How's it going?
↳ Buffering work and external file hosting site research still required.