small fix in the examples notebook

examples/notebooks/feature_overviews/0.15/new_features.ipynb

@@ -840,7 +840,11 @@
    "metadata": {},
    "cell_type": "markdown",
    "source": [
-    "40 million rows! Not bad, right? Most of this speed increase is due to Deephaven CSV's ability to parse columns directly to the target type, like `Int`, or `Double`, instead of parsing everything as a `String` and then converting it. DataFrame still reads everything into (boxed) memory, so there are limits to the size of the file you can read, but now the CSV reader is not a limiting factor anymore.\n",
+    "40 million rows! Not bad, right?\n",
+    "\n",
+    "The fact we can now read this file is due to Deephaven CSV's ability to parse columns directly to the target type, like `Int`, or `Double`, instead of reading and storing everything as a `String` first and then parsing it. This saves both memory and running time. Deephaven made a [blogpost](https://deephaven.io/blog/2022/02/23/csv-reader/) if you're curious about the specifics.\n",
+    "\n",
+    "DataFrame still reads everything into (boxed) memory, so there are limits to the size of the file you can read, but now the CSV reader is not a limiting factor anymore.\n",
     "\n",
     "Switching to the new API, in most cases, is as easy as swapping `readCSV` with `readCsv` (and `readTSV` with `readTsv`, etc.). However, there are a few differences in the API, so be sure to check the KDocs of the new functions.\n",
     "\n",
@@ -11573,7 +11577,7 @@
     "\n",
     "[Parsing](https://kotlin.github.io/dataframe/parse.html), in DataFrame, is a special case of [`convert`](https://kotlin.github.io/dataframe/convert.html).\n",
     "It can convert `String` columns to any other supported type by guessing.\n",
-    "This can be done manually, by calling `.parse()` on a dataframe, but it also happens automatically when reading from textual data, like CSV, or JSON.\n",
+    "This can be done manually, by calling `.parse()` on a dataframe, but it also happens automatically when reading from textual data, like CSV.\n",
     "\n",
     "In DataFrame 0.15:\n",
     "- The speed of parsing and guessing types has improved\n",