From 20ffd4ebf4822bbed86184dbf6ac129f268ce9bb Mon Sep 17 00:00:00 2001 From: agent Date: Thu, 5 Nov 2026 20:50:33 -0400 Subject: [PATCH] docs: the Lemonade pin moves, so stop calling the measured lemond "the server this pin ships" MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit docs/lemonade.md's HRX row said the ported `lemond` `f44ff953` is "the server this pin ships". True when written, false now: #1175 bumped third_party/lemonade to `64eb505cc62a`, so the row names a build the pin no longer carries — and this is the second time that sentence expired, which is why #1141 was closed unmerged (it said the pin was `ef0f3b49`, then #1126 landed and made the original text true again). The durable wording states the sha that was measured and points at `git ls-tree third_party/lemonade` for which one ships today, so the next bump cannot invalidate it. docs/releases.md's v2026.43 candidate table gets the same treatment in one word: "the sha `bump-lemonade`'s PR #1126 pins" -> "pinned", because that table is a dated record and the tip has moved since. Verified: site renders 57 pages rc 0 against this tree, the new sentence is in lemonade.html, one

per page, no bare closing tag added (diff grep: 0), copyright rc 0. --- docs/lemonade.md | 2 +- docs/releases.md | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/docs/lemonade.md b/docs/lemonade.md index b26d96934..a47516d21 100644 --- a/docs/lemonade.md +++ b/docs/lemonade.md @@ -121,7 +121,7 @@ What "fully embedded" still needs, measured on Strix Halo. The runs use the pinn | device | 2026-09-25, start | now | |---|---|---| -| HRX | 19 / 31 | **31 / 31 run, all pass** (2 skipped; 2026-11-05, against the **ported** `lemond` `2026.47.0~1759.f44ff953` — the server this pin ships — not the c22 package that booted the board: `test/server_llm.py --wrapped-server onebit --backend hrx`, exit status 0, 110.6 s, every backend on `HRX0`, measured at load 28.5 with two foreign HRX clients resident, engine payload `lemonade-onebit-rc-2026-11-05-c25`) | +| HRX | 19 / 31 | **31 / 31 run, all pass** (2 skipped; 2026-11-05, against the **ported** `lemond` `2026.47.0~1759.f44ff953` — the fork tip when the suite ran; the pin moves, so read `git ls-tree third_party/lemonade` for the one that ships today — not the c22 package that booted the board: `test/server_llm.py --wrapped-server onebit --backend hrx`, exit status 0, 110.6 s, every backend on `HRX0`, measured at load 28.5 with two foreign HRX clients resident, engine payload `lemonade-onebit-rc-2026-11-05-c25`) | | NPU | 3 / 31 | in progress ([npu.md](npu.md)); the suite's small test models are GGUF | | Vulkan (removed in RFC #213 stage 3) | 19 / 31 | 29 / 29 run, all pass, before its removal | diff --git a/docs/releases.md b/docs/releases.md index c24c4dfab..58bc64304 100644 --- a/docs/releases.md +++ b/docs/releases.md @@ -174,7 +174,7 @@ Release sha and pins (the submodules this candidate is built from): | engine | `7871c71d` (main when the last release-blocking PR merged; a later merge today re-runs only the stages it touches, and the run's record carries the final sha) | | llama.cpp (HRX fork) | `851fa6c8d8f2` | | hrx-system | `82804197bcbe` | -| Lemonade (fork) | `f44ff953` (the fork's `main` tip, and the sha `bump-lemonade`'s PR #1126 pins; re-read this row from `git ls-tree third_party/lemonade` at cut time, because the workflow follows the tip) | +| Lemonade (fork) | `f44ff953` (the fork's `main` tip, and the sha `bump-lemonade`'s PR #1126 pinned; re-read this row from `git ls-tree third_party/lemonade` at cut time, because the workflow follows the tip) | | private kernel add-on (HIP) | `744bd713d968` | | private NPU kernel set | `17ed5be7a28a` | | private NPU runtime | `ce9fa59adf95` |