LittleFS is an open-source, community driven project, designed and maintained by Christopher Haster (@geky).

LittleFS is by no means a finished project. There are a number of open challenges, mostly around performance and scalability, which require time to explore, solve, implement, and test.

This is my current roadmap, though it is highly susceptible to change:

1. Explore a better metadata data-structure (self-balancing Dhara trees)
   • Reduce metadata compaction O(n^2) -> O(n log(n))
   • Enable different integer sizes
2. Explore a better file data-structure (traditional B-trees)
   • Allow efficient random file writes
   • Allow logging-friendly APIs
   • Allow better block utilization for very large blocks
   • Allow data checksumming
3. Explore more robust alternatives to the dir-tree's threaded linked-list
   • The threaded linked-list has been a large source of bugs and should go
4. A global checksum system
   • Prevent "rollback" failure mode on checksum failures
   • Allow external storage of a global checksum for additional protection
5. Metadata redundancy
   • Prevent metastability issues on powerloss
   • Allow optional metadata error-correction at the block level
6. Garbage-collection APIs
   • Allow users to control when expensive garbage collection occurs
7. Optional block free-list
   • Allow more efficient block allocation
   • Allow tracking of bad blocks to avoid reuse
8. Data redundancy
   • Allow optional data error-correction at the block level
9. Different integer sizes
   • 16-bit mode => potential RAM/code savings for 16-bit devices/small storage
   • 64-bit mode => >2 GiB files, etc

Donations allow me to dedicate more of my time to LittleFS, which is split between ongoing maintenance and this long-term roadmap. These are strictly donations, no work is guaranteed, but help in making the development of LittleFS sustainable is always appreciated.

If you would like to help a different way or have other requests, feel free to reach me at geky at geky.net.