diff --git a/CHANGELOG.md b/CHANGELOG.md
index ec6ef2b..039cac6 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -1,4 +1,4 @@
-v1.2.0 (unreleased)
+v1.2.0 (2026-09-17)
 
 Fixes
 
@@ -27,7 +27,6 @@ New
 * Add `pywaffle.reload_font_awesome()`, which forgets the resolved fonts so a changed `PYWAFFLE_FONTAWESOME_DIR` takes effect without restarting
 * Add `pywaffle.font_awesome_status()`, which reports which Font Awesome is in use, where it came from, its version, how many icons each style has, and whether aliases are available. It never raises: when Font Awesome cannot be found it reports every directory searched and how to install it
 * Font Awesome can now come from the system rather than the Python package. `PYWAFFLE_FONTAWESOME_DIR` points at a directory of `.otf` files, and the usual system font directories are searched as a fallback, so a distribution's font package works on its own. Distribution packages ship fonts without Font Awesome's `icons.json`, so in that case the icon names are recovered from the fonts themselves - every canonical name is available, though aliases are not ([#25](https://github.com/gyli/PyWaffle/issues/25))
-
 * Add `rounding_rule="float"`, which draws partial blocks instead of rounding values ([#26](https://github.com/gyli/PyWaffle/issues/26)). A category that ends part way through a block fills only that fraction of it, and a block containing a boundary between two categories is split between their colors. The block count then depends only on the total of the values, so two datasets with the same total produce charts of the same size - which rounding did not guarantee
 * Add `background_color`, which fills the space behind the blocks including the gaps between them, and `block_edge_color` / `block_edge_width`, which draw a border around each block ([#37](https://github.com/gyli/PyWaffle/issues/37)). The blank cells that `block_arranging_style='new-line'` pads a line with get no border, so a padded line still ends where its value ends
 * Add `show_values` and `value_format`, which append each category's value or its percentage of the total to its legend label. This is the f-string the documentation has always told people to write by hand: `labels=[f"{k} ({v}%)" for k, v in data.items()]`
@@ -37,6 +36,7 @@ New
 
 Other
 
+* Rewrite the quickstart and the documentation homepage around a real dataset, and add a `Quickstart` page to the docs, which previously had none - it existed only in the README
 * Add opt-in image-comparison tests (`pytest tests/test_images.py --mpl`), which catch rendering regressions that block-by-block assertions cannot
 * `Waffle.make_waffle()` now raises `ValueError` when given `plots`, which it silently ignored. It draws into the single axis passed as `ax`, so `plots` never had any effect
 * Add `pywaffle.__version__`