Plain copy in the feed, share cards, page titles, newsletter, and bench

docs/bench/architecture/read-path.md

@@ -123,7 +123,7 @@ order is the curated `GROUP_ORDER` in `queries.ts`:
 Compression, Compression Size, Random Access, Clickbench,
 TPC-H (NVMe/S3) SF=1, SF=10, SF=100 (alternating),
 TPC-DS (NVMe) SF=1,
-Statistical and Population Genetics, PolarSignals Profiling,
+Statistical and Population Genetics, Polar Signals Profiling,
 fineweb (NVMe/S3), Appian (NVMe)
 ```
 

scripts/parity/README.md

@@ -120,7 +120,7 @@ JSON path with both values:
 ```text
 /api/group/{slug}?n=50: body differs $.charts[3].series["datafusion:parquet"][41]: 1.25 != 1.26
 /api/chart/{slug}?n=100: cache-control "public, s-maxage=300, stale-while-revalidate=3600" != "no-store"
-/chart/{slug}: title "TPC-H Q1 - Vortex Benchmarks" != "Vortex Benchmarks"
+/chart/{slug}: title "TPC-H Q1 | Vortex" != "Benchmarks | Vortex"
 ```
 
 When two bodies parse to the same JSON but differ in bytes, the line gives the byte offset and the

src/app/(bench)/bench/chart/[slug]/page.tsx

@@ -114,7 +114,7 @@ const getChart = cache(async (slug: string, n: string | null): Promise<ChartResp
 });
 
 /**
- * The browser-tab title mirrors v3: `{display_name} - Vortex Benchmarks`. The
+ * The browser-tab title is `{display_name} | Vortex`, the system's page title. The
  * description gives each series' newest value, and the canonical URL is the
  * chart's canonical slug without a window.
  */
@@ -133,7 +133,7 @@ export async function generateMetadata({
     return {};
   }
   const key = chartKeyFromSlug(slug);
-  const title = `${chart.display_name} - Vortex Benchmarks`;
+  const title = `${chart.display_name} | Vortex`;
   const description = chartDescription(chart, key.k);
   const url = chartUrl(chartKeyToSlug(key));
   return {

src/app/(bench)/bench/layout.tsx

@@ -20,12 +20,12 @@ import '@/bench/styles/globals.css';
 export async function generateMetadata(): Promise<Metadata> {
   const base = await requestBenchBase();
   return {
-    title: 'Vortex Benchmarks',
+    title: 'Benchmarks | Vortex',
     description: BENCH_DESCRIPTION,
     // The canonical and Open Graph URLs are bench's public ones, written out
     // in full. No `metadataBase`: it would also resolve the icons' paths, which
     // belong to the host serving the page, against bench.vortex.dev.
-    openGraph: openGraph('Vortex Benchmarks', BENCH_DESCRIPTION, `${BENCH_URL}/`),
+    openGraph: openGraph('Benchmarks | Vortex', BENCH_DESCRIPTION, `${BENCH_URL}/`),
     // Theme-aware favicons, ported from `render.rs::favicon_links`: the black
     // sigil on light-mode tabs, the white sigil on dark-mode tabs, with the dark
     // sigil as the unmediated fallback (and the apple-touch icon). The files are

src/app/(bench)/bench/methodology/page.tsx

@@ -8,7 +8,7 @@ import { BenchHeader } from '@/bench/components/BenchHeader';
 import { METHODOLOGY } from '@/bench/lib/methodology';
 import { BENCH_URL, openGraph } from '@/bench/lib/site-files';
 
-const TITLE = 'Methodology - Vortex Benchmarks';
+const TITLE = 'Methodology | Vortex';
 const DESCRIPTION =
   'How the Vortex benchmarks run: the machines, storage, engines, and formats, how each ' +
   'result is measured, and how the summaries are scored.';

src/app/(site)/blog/[slug]/opengraph-image.tsx

@@ -1,6 +1,7 @@
 import { notFound } from "next/navigation";
 import { getAllSlugs, getPostBySlug } from "@/lib/blog";
-import { OG_CONTENT_TYPE, OG_SIZE, renderPostOgImage } from "@/lib/og";
+import { siteName } from "@/lib/constants";
+import { OG_CONTENT_TYPE, OG_SIZE, renderOgCard } from "@/lib/og";
 import { formatPostDate } from "@/lib/posts";
 
 export const alt = "Vortex blog post";
@@ -22,9 +23,11 @@ export default async function OgImage({
   const post = getPostBySlug(slug);
   if (!post) notFound();
 
-  return renderPostOgImage({
-    slug,
-    title: post.title,
-    date: formatPostDate(post.date)
+  return renderOgCard({
+    seed: slug,
+    eyebrow: `${siteName} blog`,
+    title: [post.title],
+    face: "display",
+    foot: formatPostDate(post.date)
   });
 }

src/app/(site)/blog/[slug]/page.tsx

@@ -28,17 +28,17 @@ export async function generateMetadata({
 
   if (!post) {
     return {
-      title: "Post Not Found | Vortex Blog"
+      title: "Post not found | Vortex"
     };
   }
 
   const description = post.excerpt || `Read "${post.title}" on the Vortex blog`;
 
   return {
-    title: `${post.title} | Vortex Blog`,
+    title: `${post.title} | Vortex`,
     description,
     openGraph: {
-      title: `${post.title} | Vortex Blog`,
+      title: `${post.title} | Vortex`,
       description,
       siteName,
       url: `${baseUrl}/blog/${slug}`,

src/app/(site)/blog/opengraph-image.tsx

@@ -1,12 +1,16 @@
-import { OG_CONTENT_TYPE, OG_SIZE, renderOgImage } from "@/lib/og";
+import { siteName } from "@/lib/constants";
+import { OG_CONTENT_TYPE, OG_SIZE, renderOgCard } from "@/lib/og";
 
-export const alt = "Vortex Blog";
+export const alt = "Vortex blog";
 export const size = OG_SIZE;
 export const contentType = OG_CONTENT_TYPE;
 
 export default function OgImage() {
-  return renderOgImage({
-    title: "Blog",
-    description: "Updates and insights from the Vortex team."
+  return renderOgCard({
+    seed: "blog",
+    eyebrow: siteName,
+    title: ["Blog"],
+    face: "display",
+    foot: "Updates and insights from the Vortex team."
   });
 }

src/app/(site)/opengraph-image.tsx

@@ -1,13 +1,18 @@
-import { OG_CONTENT_TYPE, OG_SIZE, renderOgImage } from "@/lib/og";
+import { siteName } from "@/lib/constants";
+import { OG_CONTENT_TYPE, OG_SIZE, renderOgCard } from "@/lib/og";
 
-export const alt = "Vortex — an extensible, SOTA columnar file format";
+export const alt = "Vortex, the frontier file format.";
 export const size = OG_SIZE;
 export const contentType = OG_CONTENT_TYPE;
 
+// What Vortex is, over the homepage's standfirst, in the homepage's mono
+// voice. The card does not repeat the homepage's headline.
 export default function OgImage() {
-  return renderOgImage({
-    title: "An extensible, SOTA columnar file format",
-    description:
-      "Tools for working with compressed Apache Arrow arrays in-memory, on-disk, and over-the-wire."
+  return renderOgCard({
+    seed: "home",
+    eyebrow: siteName,
+    title: ["The frontier", "file format"],
+    face: "mono",
+    foot: "An open-source columnar file format for fast scans, random access, and compression."
   });
 }

src/app/(site)/rss/[type]/route.ts

@@ -11,15 +11,16 @@ export async function GET(
   // JSON Feed marks a link item (`external_url`). In RSS and Atom a link item
   // reads as our own post, so the three feeds would disagree.
   const posts = getAllPosts();
-  const date = new Date();
 
   const feed = new Feed({
     title: "Vortex Blog",
     description: "Updates and insights from the Vortex team",
     id: baseUrl,
     link: baseUrl,
     favicon: `${baseUrl}/favicon.ico`,
-    copyright: `All rights reserved ${date.getFullYear()}, Vortex`,
+    // The posts are CC-BY-4.0 (REUSE.toml), and the project is a Series of
+    // LF Projects, LLC.
+    copyright: "CC-BY-4.0, Vortex a Series of LF Projects, LLC",
     generator: "rss feed",
     feedLinks: {
       rss2: `${baseUrl}/rss/feed.xml`,

src/bench/components/Chart.tsx

@@ -435,7 +435,7 @@ export function Chart({ slug, name, index, groupSlug, initialPayload }: ChartIsl
             step={1}
             defaultValue={100}
             data-role="scope-slider"
-            aria-label="Custom commit window"
+            aria-label="Show commits"
             ref={sliderRef}
           />
         </div>

src/bench/components/GroupNav.test.tsx

@@ -12,7 +12,7 @@ import { GroupNav, jumpToGroup, type GroupNavItem } from '@/bench/components/Gro
 
 const GROUPS: GroupNavItem[] = [
   { name: 'Random Access', slug: 'random_access.abc' },
-  { name: 'PolarSignals Profiling', slug: 'qmg.polar' },
+  { name: 'Polar Signals Profiling', slug: 'qmg.polar' },
 ];
 
 describe('GroupNav markup', () => {
@@ -25,7 +25,7 @@ describe('GroupNav markup', () => {
     expect(html).not.toContain('group-nav-panel--open');
     expect(html).toContain('id="group-nav-panel"');
     expect(html).toContain('>Random Access</button>');
-    expect(html).toContain('>PolarSignals Profiling</button>');
+    expect(html).toContain('>Polar Signals Profiling</button>');
   });
 
   it('marks the whole menu as needing JavaScript, since neither its toggle nor its entries work without it', () => {
@@ -49,7 +49,7 @@ describe('jumpToGroup', () => {
     document.body.innerHTML = `
       <div id="group-nav-panel"></div>
       <section id="polarsignals-profiling" data-group-slug="qmg.polar">
-        <details class="group-disclosure"><summary>PolarSignals</summary></details>
+        <details class="group-disclosure"><summary>Polar Signals</summary></details>
       </section>
       <section id="random-access" data-group-slug="random_access.abc">
         <details class="group-disclosure"><summary>Random Access</summary></details>
@@ -139,7 +139,7 @@ describe('GroupNav jump-menu clicks', () => {
     (globalThis as { IS_REACT_ACT_ENVIRONMENT?: boolean }).IS_REACT_ACT_ENVIRONMENT = true;
     document.body.innerHTML = `
       <section id="polarsignals-profiling" data-group-slug="qmg.polar">
-        <details class="group-disclosure" open><summary>PolarSignals</summary></details>
+        <details class="group-disclosure" open><summary>Polar Signals</summary></details>
       </section>
       <section id="random-access" data-group-slug="random_access.abc">
         <details class="group-disclosure"><summary>Random Access</summary></details>

src/bench/components/SummaryCard.tsx

@@ -28,7 +28,7 @@ function seriesTitle(item: SeriesRanking, coverageUnit: string): string {
   const label = displaySeriesLabel(item.name);
   return item.measured >= item.total
     ? label
-    : `${label} - measured in ${item.measured} of ${item.total} ${coverageUnit}; the rest scored by the missing-series penalty`;
+    : `${label} · measured in ${item.measured} of ${item.total} ${coverageUnit} · the rest scored by the missing-series penalty`;
 }
 
 /**

src/bench/lib/anchor.test.ts

@@ -37,6 +37,10 @@ describe('groupAnchors', () => {
     ]);
   });
 
+  it('keeps the anchor a renamed group had, so shared links still land on it', () => {
+    expect(groupAnchor('Polar Signals Profiling')).toBe('polarsignals-profiling');
+  });
+
   it('keeps the full production group list collision-free', () => {
     const names = [
       'Compression',
@@ -55,7 +59,7 @@ describe('groupAnchors', () => {
       'TPC-DS (S3) (SF=100)',
       'Random Access',
       'Statistical and Population Genetics',
-      'PolarSignals Profiling',
+      'Polar Signals Profiling',
       'fineweb [nvme]',
       'fineweb [s3]',
       'appian [nvme]',

src/bench/lib/anchor.ts

@@ -11,13 +11,27 @@
  * URL. The API keeps the opaque slug.
  */
 
+/**
+ * Anchors that stay as they were when a group's name changed, so links that
+ * readers already shared still land on the group. The group was named
+ * "PolarSignals Profiling" until its owners' spelling replaced it.
+ */
+const KEPT_ANCHORS: Readonly<Record<string, string>> = {
+  'Polar Signals Profiling': 'polarsignals-profiling',
+};
+
 /**
  * Derive a URL anchor from a group display name: lowercase, every run of
  * non-alphanumerics collapsed to a single `-`, leading/trailing dashes
  * trimmed. `TPC-H (NVMe) (SF=1)` becomes `tpc-h-nvme-sf-1`. A name with no
- * alphanumerics at all falls back to `group` so the anchor is never empty.
+ * alphanumerics at all falls back to `group` so the anchor is never empty. A
+ * renamed group keeps its old anchor (`KEPT_ANCHORS`).
  */
 export function groupAnchor(name: string): string {
+  const kept = KEPT_ANCHORS[name];
+  if (kept !== undefined) {
+    return kept;
+  }
   const anchor = name
     .toLowerCase()
     .replace(/[^a-z0-9]+/g, '-')

src/bench/lib/data-cache.test.ts

@@ -69,9 +69,9 @@ describe('data-cache wrappers', () => {
     expect(BENCH_DATA_TAG).toBe('bench-data');
     expect(DATA_CACHE_BACKSTOP_SECONDS).toBe(86400);
     expect(cacheCalls.map((call) => call.keyParts)).toContainEqual([
-      'data-cache:group-charts:v10:n100',
+      'data-cache:group-charts:v11:n100',
     ]);
-    expect(cacheCalls.map((call) => call.keyParts)).toContainEqual(['data-cache:groups:v10']);
+    expect(cacheCalls.map((call) => call.keyParts)).toContainEqual(['data-cache:groups:v11']);
     for (const call of cacheCalls) {
       expect(call.options.tags).toEqual([BENCH_DATA_TAG]);
       expect(call.options.revalidate).toBe(DATA_CACHE_BACKSTOP_SECONDS);

src/bench/lib/data-cache.ts

@@ -56,7 +56,7 @@ const CACHE_OPTIONS = { tags: [BENCH_DATA_TAG], revalidate: DATA_CACHE_BACKSTOP_
  * code can serve payloads the old code computed until the next ingest.
  * It is independent of the producer-facing benchmark schema version.
  */
-const GROUP_PAYLOAD_CACHE_VERSION = 'v10';
+const GROUP_PAYLOAD_CACHE_VERSION = 'v11';
 
 // The default last-100 group bundle, keyed by group slug. The slug is the cache
 // key (an `unstable_cache` argument), so one wrapper covers every group. A

src/bench/lib/descriptions.test.ts

@@ -28,10 +28,10 @@ describe('groupDescription', () => {
         'analytics data',
     );
     expect(groupDescription('Statistical and Population Genetics')).toBe(
-      'A suite of Statistical and Population genetics queries using the gnomAD dataset',
+      'A suite of statistical and population genetics queries using the gnomAD dataset',
     );
-    expect(groupDescription('PolarSignals Profiling')).toBe(
-      'Profiling data benchmark modeled on PolarSignals/Parca, exercising scan-layer performance ' +
+    expect(groupDescription('Polar Signals Profiling')).toBe(
+      'Profiling data benchmark modeled on Polar Signals/Parca, exercising scan-layer performance ' +
         'with projection and filter pushdown on deeply nested schemas',
     );
   });

src/bench/lib/descriptions.ts

@@ -61,10 +61,10 @@ function staticDescription(name: string): string | null {
         'analytics data'
       );
     case 'Statistical and Population Genetics':
-      return 'A suite of Statistical and Population genetics queries using the gnomAD dataset';
-    case 'PolarSignals Profiling':
+      return 'A suite of statistical and population genetics queries using the gnomAD dataset';
+    case 'Polar Signals Profiling':
       return (
-        'Profiling data benchmark modeled on PolarSignals/Parca, exercising scan-layer ' +
+        'Profiling data benchmark modeled on Polar Signals/Parca, exercising scan-layer ' +
         'performance with projection and filter pushdown on deeply nested schemas'
       );
     default:

src/bench/lib/groups.test.ts

@@ -1252,16 +1252,16 @@ describe('groupNameQuery v2-name special-cases (PR-5.0.5, no DB)', () => {
     const name = groupNameQuery('statpopgen', null, null, 'nvme');
     expect(name).toBe('Statistical and Population Genetics');
     expect(groupDescription(name)).toBe(
-      'A suite of Statistical and Population genetics queries using the gnomAD dataset',
+      'A suite of statistical and population genetics queries using the gnomAD dataset',
     );
   });
 
   it('maps polarsignals to its v2 name and attaches the profiling description', () => {
     // Same null scale_factor as statpopgen in production.
     const name = groupNameQuery('polarsignals', null, null, 'nvme');
-    expect(name).toBe('PolarSignals Profiling');
+    expect(name).toBe('Polar Signals Profiling');
     expect(groupDescription(name)).toBe(
-      'Profiling data benchmark modeled on PolarSignals/Parca, exercising scan-layer ' +
+      'Profiling data benchmark modeled on Polar Signals/Parca, exercising scan-layer ' +
         'performance with projection and filter pushdown on deeply nested schemas',
     );
   });

src/bench/lib/methodology.ts

@@ -26,14 +26,14 @@ export interface MethodologySection {
 export const SCORES = {
   query:
     'the geometric mean, over the queries, of the series’ time divided by the fastest ' +
-    'series’ time on that query. 1.00x is the fastest; lower is better.',
+    'series’ time on that query. 1.00× is the fastest. Lower is better.',
   /** No vector search results reach bench yet; `llms.txt` prints this only if some do. */
   vectorSearch:
     'the geometric mean, over the thresholds, of the series’ time divided by the fastest ' +
-    'series’ time at that threshold. 1.00x is the fastest; lower is better.',
+    'series’ time at that threshold. 1.00× is the fastest. Lower is better.',
   randomAccess:
     'the geometric mean, over the datasets, of the series’ time divided by the fastest ' +
-    'series’ time on that dataset. 1.00x is the fastest; lower is better.',
+    'series’ time on that dataset. 1.00× is the fastest. Lower is better.',
   throughput:
     'the geometric mean, over the datasets, of the format’s throughput divided by ' +
     'Parquet’s. Higher is better.',
@@ -49,8 +49,8 @@ export const METHODOLOGY: readonly MethodologySection[] = [
       'Every push to the develop branch of https://github.com/vortex-data/vortex runs the ' +
         'compression, string encoding, random access, and SQL query benchmarks, and bench ' +
         'records their results. A nightly run at 00:00 UTC adds TPC-H at scale factor 100 ' +
-        'and TPC-DS at scale factors 10 and 100. Benchmarks run on pull requests only ' +
-        'comment on the pull request; their results never reach bench.',
+        'and TPC-DS at scale factors 10 and 100. Benchmarks that run on pull requests only ' +
+        'comment on the pull request. Their results never reach bench.',
     ],
   },
   {
@@ -68,7 +68,7 @@ export const METHODOLOGY: readonly MethodologySection[] = [
     text: [
       'Queries run on DataFusion 55 and DuckDB 2.0. The formats are Vortex (its default ' +
         'and compact write strategies), Apache Parquet, Lance 13, Arrow IPC, DuckDB’s own ' +
-        'format, and, for random access, LitData; not every suite runs every format. The ' +
+        'format, and, for random access, LitData. Not every suite runs every format. The ' +
         'compression and file size benchmarks write Parquet with parquet-rs 59 and zstd at ' +
         'level 3. ClickBench reads its published Parquet files.',
     ],
@@ -77,12 +77,12 @@ export const METHODOLOGY: readonly MethodologySection[] = [
     title: 'How a result is measured',
     text: [
       'SQL suites drop the operating system’s page cache before each query, run it ' +
-        'several times, and record the median time. Most suites run each query 5 times; ' +
+        'several times, and record the median time. Most suites run each query 5 times. ' +
         'TPC-H at scale factors 1 and 10 and Appian run it 10 times, TPC-DS at scale ' +
         'factor 100 runs it 3 times, and the Vortex queries run it 100 times.',
       'Random access drops the page cache before each dataset, format, access pattern, ' +
         'and open mode, repeats the lookups for 10 seconds, and records the median. Hot ' +
-        'results keep one file open after a one-second untimed warm-up; cold results open ' +
+        'results keep one file open after a one-second untimed warm-up. Cold results open ' +
         'the file again, footer included, for each lookup.',
       'The compression benchmark drops the page cache before each dataset, writes and ' +
         'scans it 5 times, and records the fastest run.',