reorder

posts/minimizer-lower-bound/minimizer-lower-bound.org

@@ -162,14 +162,14 @@ always there, ready to be found, but nobody did. Schleimers' original bound
 It really feels like this proof has been staring us in the face while we didn't
 see it for quite some time. Especially given that it's so simple, and all parts
 were hinted at:
+- The density of a forward scheme can be evaluated by considering a De Bruijn
+  sequence of order $w+k$, so looking at an order $w+k$ De Bruijn graph should
+  be necessary and sufficient.
 - Partitioning the De Bruijn graph has been done before, to show a density lower
   bound of $1/w$.
 - The density of random minimizers for $k=1$ is $2/(w+1)$, which, in hindsight,
   /very/ much reads like /at least $2$ samples are needed in every cycle of
   length $w+1$./
-- The density of a forward scheme can be evaluated by considering a De Bruijn
-  sequence of order $w+k$, so looking at an order $w+k$ De Bruijn graph should
-  be necessary and sufficient.
 
 *Story time.* Let me briefly write down how I came up with this for my own fun
 :) It's in a very very roundabout way: