Add a note on how this was accomplished for next year

Author: benjie

src/app/conf/2025/_videos.ts

@@ -1,3 +1,34 @@
+/**
+ * This was populated with data captured via:
+ *
+ * https://developers.google.com/youtube/v3/docs/playlistItems/list
+ *
+ * with:
+ *
+ * - part: snippet
+ * - maxResults: 50
+ * - playlistId: UUIVbABX6aSk5hy5UPAhQdmQ
+ *
+ * And then also adding `pageToken` for the second request to get the second
+ * page.
+ *
+ * Note that the playlistId is the ID of the "Uploads" playlist for the
+ * "GraphQLFoundationTalks" channel.
+ *
+ * Since the videos are unlisted it needed to be authenticated with the
+ * https://www.googleapis.com/auth/youtube.readonly scope. This is also
+ * why we couldn't use the generate-videos-mappings.py script.
+ *
+ * Then we simply mapped over the results:
+ *
+ * ```js
+ * for (const item of data.items) {
+ *   const id = item.snippet.resourceId.videoId;
+ *   const title = item.snippet.title
+ *   videos.push({ id, title })
+ * }
+ * ```
+ */
 export const videos: {
   id: string
   title: string