clean up

doc/batched_sequences.md

@@ -0,0 +1,26 @@
+# Batched sequences
+## Source
+You should make all your sources return `BatchSequence<T>`
+and then you can use the `.batched(count: Int)` function
+to drastically decrease the amount of single reads in the original source.
+
+Example:
+```kt
+File("text.txt")  // File
+  .openRead()     // BatchSequence<Byte>
+  .batched(64)    // BatchSequence<Byte>
+```
+
+## Sink
+You should make all your sinks take `BatchSequence<T>`
+and then you can use the `.asBatch()` function to allow
+the sink to get multiple bytes at once.
+
+Example:
+```kt
+val data = myData  // Sequence<Byte>
+  .asBatch()       // BatchSequence<Byte>
+
+File("text.txt")
+  .write(data)
+```

doc/lazy_seq.md

@@ -0,0 +1,13 @@
+# Lazy Sequences
+When writing recursive functions like Fibonacci, it is often easier and faster to use
+lazy sequences.
+
+Example:
+```kt
+val fib = lazySequence(0 to 1) { i, f ->
+  f(i-1) + f(i-2)
+}
+
+println(fib[10])
+```
+Note: If we call f for any number below 0, it will call f(0) instead.