diff --git a/TODO.md b/TODO.md
deleted file mode 100644
index b38c4f38..00000000
--- a/TODO.md
+++ /dev/null
@@ -1,99 +0,0 @@
-# Magnetism Reflectometry Reduction
-
-## From feedback:
-
-- Separate preview window for spec, off-spec, and gisans.
-- The GISANS preview should show the different plots in wl as tiles.
-
-To-do list:
-
-## Interface
-- Add log window
-- Add pop-up message boxes for errors?
-- Create a widget to change the mantid path
-- Refactor the configuration object of the NexusData class
-- Jupyter notebook as output
-- Proper mantid script as output
-- Plotting: allow for log(x)
-- Email results
-- Better reload mecanisms for previously reduced data
-- Switching between offspec and gisans causes a recalculation
-- Better progress bar reporting for off-specular and GISANS
-- Update wavelength range according to the number of points trimmed
-- Use TOF cut behind the scene so that the number of points trimmed works regardless of binning.
-- Load trimmed points while loading reduced data.
-
-## Data Manager Design
-- Add option to match direct beam cross-secion to data cross-section, otherwise sum up all cross-section in the direct beam data file.
-- Emptying cache should also delete the mantid workspaces
-
-## Reduction
-- Add / use-sangle option to output QuickNXS file.
-- Stitching  -> make Stitch1DMany take scaling factor for the first workspace
-- Add an option to choose which cross-sections to use for asymmetry calculation
-- Mantid: add option not to rebin in Q at the end.
-- Write output algo that takes group workspace.
-- Automate selection of const-Q binning
-- Add TOF range selector
-- Compute proper resolution for const-Q binning
-
-## Off-spec
-- Only recalculate off-spec when needed (need to add a need_recalc data member)
-- Off-spec needs to use option to skip first and last points when displaying the preview
-- Try using events directly when computing offspecular.
-- Plot slices in 2D off-specular data
-
-## GISANS
-- Add background subtraction
-- Use multi-processes
-
-## Testing
-- Automated testing
-- Test numpy output
-- Test Matlab output
-
-## DONE
-- Generate GISANS output
-- Add Q binning option for output
-- Check that the overall normalization is the same as with old QuickNXS when validating
-- Add off-spec coordinates to toolbar
-- Verify off-spec normalization vs peak ranges and two-theta
-- Use PolarizerLabel and AnalyzerLabel to better decide on cross-section names
-- Use Analyzer log to decide if the analyzer was used and name cross-sections accordingly.
-- Update affected reflectivity when a direct beam run is deleted or changed.
-- Make sure all the config properties make it to QSettings
-- Add mantid script as a possible output file
-- Background subtraction for offspec
-- Trim workspaces when stitching
-- Add option for constant Q
-- Do something with DIRPIX and DANGLE0 overwrite
-- Make sure matplotlib widgets work.
-- Improve data manager design to handle direct beams more cleanly
-- SORT RUNS WHEN LOADING REDUCED FILE
-- Add progress bar
-- Normalize output to 1 (broken when turned off).
-- Norm TOF-lambda doesn't work.
-- Do something with P0 / PN to clean up the reflectivity
-- Add indicator for whether the application things the data set we are looking at is a direct beam or not.
-- Decide whether to keep peak ranges common to all cross-sections.
-- Switch active data when clicking on the reduction table.
-- Tie main UI to normalization list as well as reduction list.
-- When clicking on an entry in the reduction table:
- - Flag as active with a color indicator.
- - Get it's config and push it to the main UI.
-
-- When loading a data set from cache:
- - Check whether it's also in the reduction table.
- - If so, get it's config and push it to the main UI.
-
-- When updating the main UI, check whether we need to update the reduction table.
-- Match direct beam should update the reduction list.
-- Implement item handling in the direct beam list to be identical to the scattering run list.
-- Calculated sangle needs to update if we change the position ranges
-- Keep the Mantid workspaces instead of deleting them so we can recalculate faster.
-- Add filter to direct beam data too.
-- Allow for direct beam data to be a workspace instead of just a run number.
-- Add InputNormalizationWorkspace as input to MagnetismReflectometerReduction
-- Fixed saving/loading of peak ranges
-- Add option to cut the first N and last M points -> Needs to be saved properly to file
-- Deal with peak position when it's not in the middle of the range