fix: home page code snippets cuasing various problems

.github/workflows/docs-deploy.yml

@@ -59,27 +59,18 @@ jobs:
             - Namespace Spec: namespace
           EOF
           cp docs/src/format/namespace/rest.yaml docs/src/rest.yaml
-          cp lance-namespace/CONTRIBUTING.md docs/src/community/contributing/namespace.md
-          # Update contributing .pages to include namespace.md in navbar
-          echo "  - Namespace: namespace.md" >> docs/src/community/contributing/.pages
       - name: Copy lance-spark docs
         run: |
           cp -r lance-spark/docs/src docs/src/integrations/spark
           cat >> docs/src/integrations/.pages << 'EOF'
             - Apache Spark: spark
           EOF
-          cp lance-spark/CONTRIBUTING.md docs/src/community/contributing/spark.md
-          # Update contributing .pages to include spark.md in navbar
-          echo "  - Apache Spark: spark.md" >> docs/src/community/contributing/.pages
       - name: Copy lance-ray docs
         run: |
           cp -r lance-ray/docs/src docs/src/integrations/ray
           cat >> docs/src/integrations/.pages << 'EOF'
             - Ray: ray
           EOF
-          cp lance-ray/CONTRIBUTING.md docs/src/community/contributing/ray.md
-          # Update contributing .pages to include ray.md in navbar
-          echo "  - Ray: ray.md" >> docs/src/community/contributing/.pages
       - name: Deploy
         working-directory: docs
         run: uv run mkdocs gh-deploy --force

docs/src/community/contributing/index.md → CONTRIBUTING.md

@@ -38,7 +38,7 @@ Currently Lance is implemented in Rust and comes with a Python wrapper. So you'l
 
 ## Example Notebooks
 
-Example notebooks are under `examples`. 
+Example notebooks are under `examples`.
 These are standalone notebooks you should be able to download and run.
 
 ## Benchmarks
@@ -50,5 +50,5 @@ These live under `benchmarks`.
 
 ## Code of Conduct
 
-We follow the Code of Conduct of [Python Foundation](https://www.python.org/psf/conduct/) and 
-[Rust Foundation](https://www.rust-lang.org/policies/code-of-conduct). 
+We follow the Code of Conduct of [Python Foundation](https://www.python.org/psf/conduct/) and
+[Rust Foundation](https://www.rust-lang.org/policies/code-of-conduct).

docs/src/community/contributing/docs.md → docs/CONTRIBUTING.md

@@ -29,4 +29,4 @@ as a part of the release process.
 ### Java Generated Doc
 
 Java code documentation is built and published to Maven Central.
-You can find the doc page for the specific project at [javadoc.io](https://javadoc.io).
+You can find the doc page for the specific project at [javadoc.io](https://javadoc.io).

docs/mkdocs.yml

@@ -13,14 +13,14 @@ theme:
   favicon: logo/logo.png
   palette:
     - scheme: default
-      primary: indigo
-      accent: indigo
+      primary: custom
+      accent: custom
       toggle:
         icon: material/brightness-7
         name: Switch to dark mode
     - scheme: slate
-      primary: indigo
-      accent: indigo
+      primary: custom
+      accent: custom
       toggle:
         icon: material/brightness-4
         name: Switch to light mode
@@ -67,13 +67,12 @@ plugins:
       proto_dir: ../protos
 
 extra:
+  generator: false
   social:
     - icon: fontawesome/brands/github
       link: https://github.com/lancedb/lance
     - icon: fontawesome/brands/discord
       link: https://discord.gg/zMM32dvNtd
-    - icon: fontawesome/brands/twitter
-      link: https://twitter.com/lancedb
 
 copyright: © 2025 Lance Format. All rights reserved.
 

docs/overrides/home.html

@@ -36,8 +36,84 @@
     .md-main__inner {
       margin: 0;
     }
+
+    /* Style Learn More and View Integrations links */
+    .lance-feature-section .md-button,
+    .lance-intro-section .md-button:not(.md-button--primary) {
+      background-color: transparent;
+      color: #625EFF;
+      text-decoration: underline;
+      border: none;
+      box-shadow: none;
+    }
+
+    .lance-feature-section .md-button:hover,
+    .lance-intro-section .md-button:not(.md-button--primary):hover {
+      background-color: transparent;
+      color: #757575;
+      text-decoration: underline;
+    }
   </style>
 
+  <script>
+    // Fix palette toggle jumping to bottom of page
+    (function() {
+      // Store scroll position
+      let savedScrollY = 0;
+      let isTogglingPalette = false;
+
+      // Prevent hash-based scrolling for palette changes
+      if ('scrollRestoration' in history) {
+        history.scrollRestoration = 'manual';
+      }
+
+      // Intercept scroll attempts during palette toggle
+      const preventScroll = function() {
+        if (isTogglingPalette) {
+          window.scrollTo(0, savedScrollY);
+        }
+      };
+
+      // Save scroll position and intercept clicks on palette toggle
+      document.addEventListener('click', function(e) {
+        const paletteLabel = e.target.closest('[for^="__palette"]');
+        if (paletteLabel) {
+          savedScrollY = window.scrollY;
+          isTogglingPalette = true;
+
+          // Prevent any scrolling for a brief moment
+          window.addEventListener('scroll', preventScroll, { passive: false });
+
+          setTimeout(function() {
+            isTogglingPalette = false;
+            window.removeEventListener('scroll', preventScroll);
+          }, 100);
+        }
+      }, true);
+
+      // Listen for hash changes from palette toggle
+      window.addEventListener('hashchange', function(e) {
+        if (window.location.hash.startsWith('#__palette')) {
+          e.preventDefault();
+          // Immediately restore scroll position
+          window.scrollTo(0, savedScrollY);
+          // Clean up hash without triggering another hashchange
+          history.replaceState(null, null, window.location.pathname);
+        }
+      }, false);
+
+      // Also handle palette input changes directly
+      document.addEventListener('DOMContentLoaded', function() {
+        const paletteInputs = document.querySelectorAll('input[name="__palette"]');
+        paletteInputs.forEach(function(input) {
+          input.addEventListener('change', function() {
+            window.scrollTo(0, savedScrollY);
+          });
+        });
+      });
+    })();
+  </script>
+
   <!-- Hero Section -->
   <section class="mdx-container">
     <div class="container">
@@ -76,7 +152,7 @@ <h2>What is Lance<sup>™</sup>?</h2>
           allowing you to build a complete open lakehouse on top of object storage to power your AI workflows.
           Lance brings high-performance vector search, full-text search, random access, and feature
           engineering capabilities to the lakehouse, while you can still get all the existing lakehouse benefits
-          like SQL analytics, ACID transactions, time travel, and integrations with open engines (Apache Spark, Ray, Trino, DuckDB, etc.)
+          like SQL analytics, ACID transactions, time travel, and integrations with open engines (Apache Spark, Ray, PyTorch, Trino, DuckDB, etc.)
           and open catalogs (Apache Polaris, Unity Catalog, Apache Gravitino, Hive Metastore, etc.)
         </p>
         <a href="quickstart" class="md-button md-button--primary">Learn More</a>
@@ -93,7 +169,7 @@ <h2>Expressive Hybrid Search</h2>
           <p>
             Lance enables powerful hybrid search combining vector similarity, full-text search,
             and SQL analytics on the same dataset. All query types are accelerated by corresponding
-            secondary indices as part of the Lance specification.
+            secondary indexes as part of the Lance specification.
           </p>
           <p>
             Run semantic search on embeddings, BM25 search on keywords, and apply complex SQL predicates -
@@ -102,19 +178,7 @@ <h2>Expressive Hybrid Search</h2>
           <a href="quickstart/vector-search" class="md-button">Learn More</a>
         </div>
         <div class="lance-feature-demo">
-          <div id="termynal-hybrid-search"
-            data-termynal=""
-            data-ty-startdelay="500"
-            data-ty-typedelay="40"
-            data-ty-linedelay="700"
-            style="width: 500px;">
-            <span data-ty="input" data-ty-prompt=">>>">import lance</span>
-            <span data-ty="input" data-ty-prompt=">>>">ds = lance.dataset("s3://my-bucket/docs")</span>
-            <span data-ty="input" data-ty-prompt=">>>">ds.to_table(full_text_query="machine learning")</span>
-            <span data-ty="input" data-ty-prompt=">>>">ds.to_table(</span>
-            <span data-ty="input" data-ty-prompt="...">  nearest={"column": "embedding", "q": query_vec, "k": 10},</span>
-            <span data-ty="input" data-ty-prompt="...">  filter="year > 2020")</span>
-          </div>
+          <img src="assets/images/hybrid-search.png" alt="Hybrid Search Example" style="max-width: 500px; width: 100%; height: auto; border-radius: 8px;">
         </div>
       </div>
     </div>
@@ -127,28 +191,19 @@ <h2>Expressive Hybrid Search</h2>
         <div class="lance-feature-text">
           <h2>Lightning-fast Random Access</h2>
           <p>
-            Lance delivers 100x faster random access compared to Parquet or Iceberg. With efficient
-            row-addressing, you can access individual records across multiple files instantly,
-            making it perfect for real-time ML serving, random sampling, and interactive applications.
+            Lance delivers 100x faster random access compared to Parquet or Iceberg.
+            Unlike traditional formats, Lance maintains high performance even when
+            randomly accessing scattered rows across your entire dataset.
           </p>
           <p>
-            Unlike traditional columnar formats, Lance maintains high performance even when
-            randomly accessing scattered rows across your entire dataset.
+            With a highly optimized file format plus efficient row-addressing and secondary indexes at table level,
+            you can access individual records across multiple files instantly,
+            making it perfect for real-time ML serving, random sampling, and interactive applications.
           </p>
           <a href="guide/read_and_write#random-access" class="md-button">Learn More</a>
         </div>
         <div class="lance-feature-demo">
-          <div id="termynal-random-access"
-            data-termynal=""
-            data-ty-startdelay="1500"
-            data-ty-typedelay="40"
-            data-ty-linedelay="700"
-            style="width: 500px;">
-            <span data-ty="input" data-ty-prompt=">>>">import lance</span>
-            <span data-ty="input" data-ty-prompt=">>>">dataset = lance.dataset("s3://my-bucket/embeddings")</span>
-            <span data-ty="input" data-ty-prompt=">>>">table = dataset.take([100, 5000, 1000000])</span>
-            <span data-ty="input" data-ty-prompt=">>>">dataset.take([0, 1], columns=["id", "vector"])</span>
-          </div>
+          <img src="assets/images/random-access.png" alt="Random Access Example" style="max-width: 500px; width: 100%; height: auto; border-radius: 8px;">
         </div>
       </div>
     </div>
@@ -161,31 +216,18 @@ <h2>Lightning-fast Random Access</h2>
         <div class="lance-feature-text">
           <h2>Native Multimodal Data Support</h2>
           <p>
-            Store images, videos, audio, text, and embeddings in a single unified format.
+            Store images, videos, audio, text, and embeddings alongside your traditional tabular data in a single unified format.
             Lance's blob encoding efficiently handles large binary objects with lazy loading,
             while optimized vector storage accelerates similarity search.
           </p>
           <p>
-            Perfect for AI/ML workloads where you need to store raw data alongside embeddings
-            for multimodal retrieval and generation workflows.
+            Perfect for AI/ML workloads where you need to store raw data, ML features, generated captions and embeddings
+            all together for multimodal retrieval and genAI workflows.
           </p>
           <a href="guide/blob" class="md-button">Learn More</a>
         </div>
         <div class="lance-feature-demo">
-          <div id="termynal-multimodal"
-            data-termynal=""
-            data-ty-startdelay="2500"
-            data-ty-typedelay="40"
-            data-ty-linedelay="700"
-            style="width: 500px;">
-            <span data-ty="input" data-ty-prompt=">>>">import lance</span>
-            <span data-ty="input" data-ty-prompt=">>>">import pyarrow as pa</span>
-            <span data-ty="input" data-ty-prompt=">>>">schema = pa.schema([</span>
-            <span data-ty="input" data-ty-prompt="...">  pa.field("video", pa.large_binary(),</span>
-            <span data-ty="input" data-ty-prompt="...">    metadata={"lance-encoding:blob": "true"}),</span>
-            <span data-ty="input" data-ty-prompt="...">  pa.field("embedding", pa.list_(pa.float32(), 128))])</span>
-            <span data-ty="input" data-ty-prompt=">>>">lance.write_dataset(table, "multimodal.lance", schema=schema)</span>
-          </div>
+          <img src="assets/images/multimodal-data.png" alt="Multimodal Data Example" style="max-width: 500px; width: 100%; height: auto; border-radius: 8px;">
         </div>
       </div>
     </div>
@@ -200,7 +242,7 @@ <h2>Data Evolution > Schema Evolution</h2>
           <p>
             Schema evolution in most open table formats are metadata only and fast.
             But when trying to backfill column values in existing rows, a full table rewrite is typically required.
-            Lance supports efficient schema evolution with backfill, making it perfect for ML
+            Lance supports data evolution (efficient schema evolution with backfill), making it perfect for ML
             feature engineering, embedding and media content management.
           </p>
           <p>
@@ -210,20 +252,7 @@ <h2>Data Evolution > Schema Evolution</h2>
           <a href="guide/data_evolution" class="md-button">Learn More</a>
         </div>
         <div class="lance-feature-demo">
-          <div id="termynal-evolution"
-            data-termynal=""
-            data-ty-startdelay="3500"
-            data-ty-typedelay="40"
-            data-ty-linedelay="700"
-            style="width: 500px;">
-            <span data-ty="input" data-ty-prompt=">>>">import lance</span>
-            <span data-ty="input" data-ty-prompt=">>>">dataset = lance.dataset("my_data.lance")</span>
-            <span data-ty="input" data-ty-prompt=">>>">@lance.batch_udf()</span>
-            <span data-ty="input" data-ty-prompt="...">def add_embeddings(batch):</span>
-            <span data-ty="input" data-ty-prompt="...">  vectors = model.encode(batch["text"])</span>
-            <span data-ty="input" data-ty-prompt="...">  return {"embedding": vectors}</span>
-            <span data-ty="input" data-ty-prompt=">>>">dataset.add_columns(add_embeddings)</span>
-          </div>
+          <img src="assets/images/data-evolution.png" alt="Data Evolution Example" style="max-width: 500px; width: 100%; height: auto; border-radius: 8px;">
         </div>
       </div>
     </div>
@@ -237,11 +266,13 @@ <h2>Data Evolution > Schema Evolution</h2>
           <h2>Rich Ecosystem Integrations</h2>
           <p>
             As an open format, Lance integrates seamlessly with the Python data ecosystem and modern data platforms.
-            Work with your favorite tools including Pandas, Polars, and PyTorch for data processing and machine learning.
+            Work with your favorite tools including Pandas, Polars, Ray and PyTorch for data processing and machine learning.
+          </p>
+          <p>
             Connect with leading query engines like Apache DataFusion, DuckDB, Apache Spark, Trino, and Apache Flink/Fluss
             to run SQL analytics and distributed processing on your Lance datasets.
           </p>
-          <a href="integrations/duckdb" class="md-button">View Integrations</a>
+          <a href="integrations/datafusion" class="md-button">View Integrations</a>
         </div>
         <div class="lance-feature-demo">
           <img src="assets/images/ecosystem-integrations.png" alt="Lance Ecosystem Integrations" style="max-width: 500px; width: 100%; height: auto; border-radius: 8px;">
@@ -250,10 +281,6 @@ <h2>Rich Ecosystem Integrations</h2>
     </div>
   </section>
 
-  <script
-    src="assets/javascript/termynal.js"
-    data-termynal-container="#termynal-hybrid-search|#termynal-random-access|#termynal-multimodal|#termynal-evolution">
-  </script>
 
 {% endblock %}
 

docs/src/assets/stylesheets/home.css

@@ -1,5 +1,14 @@
 /* Lance Homepage Styles */
 
+/* Override with custom color #625EFF site-wide */
+:root > * {
+  --md-primary-fg-color: #625EFF;
+  --md-primary-fg-color--light: #8481FF;
+  --md-primary-fg-color--dark: #4A46CC;
+  --md-accent-fg-color: #625EFF;
+  --md-accent-fg-color--transparent: rgba(98, 94, 255, 0.1);
+}
+
 * {
   box-sizing: border-box;
 }
@@ -108,7 +117,7 @@
 }
 
 .intro-message .md-button--primary:hover {
-  box-shadow: 0 8px 20px rgba(102, 126, 234, 0.5);
+  box-shadow: 0 8px 20px rgba(98, 94, 255, 0.5);
 }
 
 .intro-message .md-button:not(.md-button--primary):hover {

docs/src/community/.pages

@@ -1,3 +1,2 @@
 nav:
-  - Contributing: contributing
   - Lancelot Program: lancelot.md

docs/src/community/contributing/python.md → python/CONTRIBUTING.md

@@ -29,4 +29,4 @@ make integtest
 ```
 
 To run the tests on OS X, you may need to increase the default limit on the number of open files:
-`ulimit -n 2048`
+`ulimit -n 2048`

docs/src/community/contributing/rust.md → rust/CONTRIBUTING.md

@@ -33,8 +33,8 @@ If you're working on a performance related feature, benchmarks can be run via:
 cargo bench
 ```
 
-If you want detailed logging and full backtraces, set the following environment variables. 
-More details can be found [here](../../guide/performance.md#logging).
+If you want detailed logging and full backtraces, set the following environment variables.
+More details can be found [here](../docs/src/guide/performance.md#logging).
 
 ```bash
 LANCE_LOG=info RUST_BACKTRACE=FULL <cargo-commands>