{"messages":[{"content":"You are a constitutional council ranking individual git commits for ownership allocation.\n\nCompare these two commits. Decide which contributed more lasting value to the project.\n\nJudge substance, not spectacle:\n- Prefer correct, lasting design and real bugfixes over churn, formatting, renames, or generated noise.\n- Prefer clarity and necessity over sheer line count. A small precise change can beat a large diffuse one.\n- Do not favor a side merely because its patch is longer or noisier.\n- Weight what the change does for the project, not the contributor's name.\n\nReturn ONLY a JSON object: {\"winner\": \"A\" or \"B\", \"ratio\": \"N:M\", \"explanation\": \"...\"}\nThe explanation must cite concrete differences in the patches (1-3 sentences).\n\nSide A — contributor: tommy-mor\nSide A — commit message:\n[0366806e] Color rank rows by vote mass within each group, not list position.\n\nMin–max normalization keeps similar scores visually close while still\nusing the full gradient as groups grow and absolute mass dilutes.\n\nCo-authored-by: Cursor <cursoragent@cursor.com>\n\nSide A — unified diff (full patch):\ndiff --git a/server/src/html/mod.rs b/server/src/html/mod.rs\nindex a58cbbee3490a08a625cb06df06848c59a615d65..4eff2e19ed4d303ff8e80c1eabd8a15b4990e643 100644\n--- a/server/src/html/mod.rs\n+++ b/server/src/html/mod.rs\n@@ -264,12 +264,19 @@ pub fn scope_theme_style(parent: &ItemId) -> String {\n     )\n }\n \n-fn rank_row_style(parent: &ItemId, ordinal: usize, total: usize) -> String {\n-    let t = if total <= 1 {\n-        0.0\n-    } else {\n-        ordinal as f64 / (total - 1) as f64\n-    };\n+/// Map vote mass to gradient position using the group's score range, not raw mass or\n+/// list position. Vote mass sums to 1 across the component, so absolute values dilute\n+/// as N grows; min–max within the visible list preserves similar scores → similar colors.\n+fn score_gradient_t(score: f64, min_score: f64, max_score: f64) -> f64 {\n+    let spread = max_score - min_score;\n+    if spread < 1e-9 {\n+        return 0.5;\n+    }\n+    ((max_score - score) / spread).clamp(0.0, 1.0)\n+}\n+\n+fn rank_row_style(parent: &ItemId, score: f64, min_score: f64, max_score: f64) -> String {\n+    let t = score_gradient_t(score, min_score, max_score);\n     let base_hue = scope_base_hue(parent);\n     let hue = (base_hue + 118.0 * t) % 360.0;\n     let lightness = 0.74 - 0.34 * t;\n@@ -295,14 +302,15 @@ fn rank_list(\n     highlighted: &HashSet<ItemId>,\n     tree: &GlobalTree,\n ) -> Markup {\n-    let group_len = items.len();\n+    let min_score = items.iter().map(|r| r.score).fold(f64::INFINITY, f64::min);\n+    let max_score = items.iter().map(|r| r.score).fold(f64::NEG_INFINITY, f64::max);\n     html! {\n         @if !items.is_empty() {\n             h3 class=\"rank-heading muted small\" { (label) }\n             ol class=\"rank-list\" {\n                 @for (i, r) in items.iter().enumerate() {\n                     @let href = item_href(&r.item);\n-                    @let style = rank_row_style(parent, i, group_len);\n+                    @let style = rank_row_style(parent, r.score, min_score, max_score);\n                     @let class = rank_row_class(&r.item, highlighted);\n                     li class=(class)\n                         data-rank-item=(r.item.as_str())\n@@ -517,19 +525,37 @@ pub async fn browse(State(state): State<AppState>, uri: Uri) -> impl IntoRespons\n \n #[cfg(test)]\n mod tests {\n-    use super::{rank_row_style, SORTER_UI_JS};\n+    use super::{rank_row_style, score_gradient_t, SORTER_UI_JS};\n     use crate::path_types::ItemId;\n \n     #[test]\n-    fn rank_row_style_gradients_per_group_not_globally() {\n+    fn score_gradient_t_uses_group_range_not_absolute_mass() {\n+        assert!((score_gradient_t(0.12, 0.08, 0.12) - 0.0).abs() < 1e-9);\n+        assert!((score_gradient_t(0.08, 0.08, 0.12) - 1.0).abs() < 1e-9);\n+        // Raw 12% mass would map near the dark end globally; within this group it's the top.\n+        assert!(score_gradient_t(0.12, 0.08, 0.12) < score_gradient_t(0.12, 0.0, 1.0));\n+    }\n+\n+    #[test]\n+    fn score_gradient_t_similar_scores_similar_t() {\n+        let a = score_gradient_t(0.41, 0.20, 0.60);\n+        let b = score_gradient_t(0.40, 0.20, 0.60);\n+        assert!((a - b).abs() < 0.05);\n+        assert!((a - score_gradient_t(0.60, 0.20, 0.60)).abs() > 0.3);\n+    }\n+\n+    #[test]\n+    fn score_gradient_t_tied_scores_neutral() {\n+        assert!((score_gradient_t(0.25, 0.25, 0.25) - 0.5).abs() < 1e-9);\n+    }\n+\n+    #[test]\n+    fn rank_row_style_same_inputs_same_color() {\n         let parent = ItemId::opaque(\"test-scope\");\n-        let first_in_four = rank_row_style(&parent, 0, 4);\n-        let last_in_four = rank_row_style(&parent, 3, 4);\n-        let first_in_two = rank_row_style(&parent, 0, 2);\n-        let last_in_two = rank_row_style(&parent, 1, 2);\n-        assert_eq!(first_in_four, first_in_two);\n-        assert_eq!(last_in_four, last_in_two);\n-        assert_ne!(first_in_four, last_in_four);\n+        assert_eq!(\n+            rank_row_style(&parent, 0.33, 0.20, 0.60),\n+            rank_row_style(&parent, 0.33, 0.20, 0.60),\n+        );\n     }\n \n     #[test]\n\n\nSide B — contributor: tommy-mor\nSide B — commit message:\n[047b82bd] Remove redundant zero-ratio guard from reducer.\n\nZeros are already rejected at the DSL parser and browser handler;\nthe guard in apply_vote was dead code. The negative clamping stays\nsince add_edge_weight already skips weight-0 edges correctly.\n\nCo-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>\n\nSide B — unified diff (full patch):\ndiff --git a/server/src/reducer.rs b/server/src/reducer.rs\nindex 0e36979abe0f051493038ff7e652efc7f7a0ac80..efbf7f67ff24c7a2989102fe62879b2794a27c5f 100644\n--- a/server/src/reducer.rs\n+++ b/server/src/reducer.rs\n@@ -112,11 +112,6 @@ impl GroupState {\n         if vote.ratio_right < 0 {\n             vote.ratio_right = 0;\n         }\n-        if vote.ratio_left == 0 || vote.ratio_right == 0 {\n-            // Zero on either side produces no valid edge; drop before registering items or pair.\n-            return;\n-        }\n-\n         let a_idx = self.ensure_item(&vote.a);\n         let b_idx = self.ensure_item(&vote.b);\n \ndiff --git a/server/tests/basic.rs b/server/tests/basic.rs\nindex 08159f4a7f0850fd165817a4a1af4f31ced2ad76..1748769ccf7196b2d81cf556d5368c7e723f6a4e 100644\n--- a/server/tests/basic.rs\n+++ b/server/tests/basic.rs\n@@ -533,7 +533,7 @@ fn dsl_parse_rejects_zero_zero_vote_ratio() {\n #[test]\n fn reducer_negative_ratio_clamped_to_zero() {\n     let _state = ReducerState::default();\n-    // GroupState::apply_vote clamps negatives to 0; when either side is 0 the vote is dropped.\n+    // apply_vote clamps negatives to 0; add_edge_weight skips zero-weight edges.\n     let mut group = GroupState::new();\n     group.apply_vote(slugsocial_server::reducer::VoteData {\n         ts: 1,\n@@ -546,10 +546,9 @@ fn reducer_negative_ratio_clamped_to_zero() {\n         delegate: Some(\"00000000-0000-0000-0000-000000000000:test:local/test\".to_string()),\n         thread_tag: \"t\".to_string(),\n     });\n-    // Nothing registered: zero-clamped vote is dropped before ensure_item.\n-    assert!(group.idx_to_item.is_empty());\n+    // Items and pair are registered; edges are absent because weight 0 is skipped.\n+    assert_eq!(group.idx_to_item.len(), 2);\n     assert!(group.edges.is_empty());\n-    assert!(group.voted_pairs.is_empty());\n }\n \n \n","role":"user"}],"model":"openai/gpt-chat-latest"}