build(release): 7 more contributed pipeline lists from #14, 555 pipelines #2041

Merged
HumanBean merged 1 commit from HumanBean/Longshot:build/pipelines-from-14-2 into main 2026-11-21 07:53:20 +00:00
Owner

Summary

Since #1688, nine more players have attached the pipeline lists their game wrote to the Submit your pipelines issue. This PR adds the seven that hold something new. The first start now prepares 555 more pipelines before the title: 76,567 in all, up from 76,012. That should mean fewer stutters, or white washes, the first time a player reaches an area. It adds the most for AMD with FSR 4 (Maver1ck's RX 7900 XTX at night in Kakariko, 337 new), DLSS on Linux (bwen and jlh4426), and DLAA with frame generation (Muramasa82).

The files to read are the seven assets/pipeline_states/contrib/*.md notes. Nothing else changes: no code, no other list, and the merge order is the same.

List Who GPU and driver New pipelines
arturito-604-rtx-4060 Arturito-604 NVIDIA GeForce RTX 4060 8 GB, driver NVIDIA 621.24, Windows 11 9
bwen-rtx-3060-linux bwen NVIDIA GeForce RTX 3060, driver NVIDIA 595.99.02, Ubuntu 22.04 94
icastot-rx-6750-xt icastot AMD Radeon RX 6750 XT, AMD 26.11.2, Windows 11 26
jlh4426-rtx-5060-linux jlh4426 NVIDIA GeForce RTX 5060, driver NVIDIA 620.70.06, PikaOS 4 11
maver1ck-rx-7900-xtx Maver1ck AMD Radeon RX 7900 XTX, AMD 26.11.1 (LLPC), Windows 11 337
muramasa82-rtx-4070-super Muramasa82 NVIDIA GeForce RTX 4070 SUPER, driver NVIDIA 621.24, Windows 11 77
sx-nn-rtx-4070-super sx.nn NVIDIA GeForce RTX 4070 SUPER, driver NVIDIA 610.88, Windows 11 1

Not added: MB44's list (292 pipelines) and piano_butcher's (142). Everything in them is already in other lists, as #1688 also found.

How the zips were handled:

  • Each attachment was downloaded into a folder of its own, and its members were listed.
  • All nine held only longshot_pipeline_states.bin, and eight also held longshot_pipeline_report.txt. Only those members were read, and nothing was run.
  • Each list passed nx2/pipeline_states.py --check before add. All are version 2.
  • The notes' GPU, build and settings come from the zip's report, or from the comment for Arturito-604, whose zip has no report. The route comes from the comment, or "Not given" where the player left it blank.

See #14 (it stays open for more lists).

Testing

Tested: nx2/pipeline_states.py --check passed on each of the nine sent lists. After the adds, nx2/pipeline_states.py check passes on the sources: merged: 76567 pipelines (compute 3357, graphics 68877, mesh 4333) of 27276 programs, version 2. That is under COMPUTE_MOST (3584). The test suite was not run on this PC. There was no game run: the PR was opened with --force at the maintainer's request, because it changes no game code, only the data the first start prepares.

Still to test:

  • Any GPU, a fresh game folder built from this branch: let the first start's shader preparation run to the title, then play a few minutes. Expect it to finish without errors, and the log to report 555 more pipelines prepared than on main (by hand).

Measurements

Verdict: nothing measured in the game; the data only adds 555 pipelines the first start prepares. The counts below come from nx2/pipeline_states.py check.

Before: main @ 7c23d2be After: this PR @ d68cd654
merged pipelines 76,012 76,567
graphics 68,393 68,877
mesh 4,310 4,333
compute programs 3,309 3,357
programs 27,125 27,276

Risks

  • First start: preparation takes a little longer, for 555 more pipelines (under 1 % more).
  • Memory: the 48 more compute programs are kept ready on every start. By the measurement in nx2/pipeline_states.py, that is about 1 MB and a few ms each, so roughly 50 MB and well under a second per start.
  • Turning it off: revert the commit.
  • After updating: a player's first start prepares the new pipelines once. Saves, options and caches are otherwise unchanged.

Checklist

Ready for review

  • One topic, on current main; the title is a Conventional Commit.
  • Tested on my own PC, in the game for a change the game runs, as Testing says. (Not run in the game: data only, opened with --force at the maintainer's request, as #1688.)
  • mise run test passes (any failure that also fails on main is named). (Not run on this PC; only nx2/pipeline_states.py check, which passes.)
  • Measured against main, or Measurements says why there is nothing to measure.
  • No game files, build output or hand-edited generated files; third_party/ untouched; nothing GPL-incompatible.

When it applies

  • AI assistance recorded with Assisted-by: / Co-Authored-By: trailers.

Ready to merge

  • Everything in Still to test has run, its results in Tested.
  • Tier 1 hardware it can affect is tested (docs/dev/standards.md, "Hardware tiers").
  • Approved in review, every comment answered.
  • CI passes on the head; on current main, measured again or a comment says why the numbers still hold.

🤖 Generated with Claude Code

## Summary Since [#1688](https://git.magicbean.dev/MagicBeanDev/Longshot/pulls/1688), nine more players have attached the pipeline lists their game wrote to the [Submit your pipelines](https://git.magicbean.dev/MagicBeanDev/Longshot/issues/14) issue. This PR adds the seven that hold something new. The first start now prepares 555 more pipelines before the title: 76,567 in all, up from 76,012. That should mean fewer stutters, or white washes, the first time a player reaches an area. It adds the most for AMD with FSR 4 (Maver1ck's RX 7900 XTX at night in Kakariko, 337 new), DLSS on Linux (bwen and jlh4426), and DLAA with frame generation (Muramasa82). The files to read are the seven `assets/pipeline_states/contrib/*.md` notes. Nothing else changes: no code, no other list, and the merge order is the same. | List | Who | GPU and driver | New pipelines | |---|---|---|---| | `arturito-604-rtx-4060` | Arturito-604 | NVIDIA GeForce RTX 4060 8 GB, driver NVIDIA 621.24, Windows 11 | 9 | | `bwen-rtx-3060-linux` | bwen | NVIDIA GeForce RTX 3060, driver NVIDIA 595.99.02, Ubuntu 22.04 | 94 | | `icastot-rx-6750-xt` | icastot | AMD Radeon RX 6750 XT, AMD 26.11.2, Windows 11 | 26 | | `jlh4426-rtx-5060-linux` | jlh4426 | NVIDIA GeForce RTX 5060, driver NVIDIA 620.70.06, PikaOS 4 | 11 | | `maver1ck-rx-7900-xtx` | Maver1ck | AMD Radeon RX 7900 XTX, AMD 26.11.1 (LLPC), Windows 11 | 337 | | `muramasa82-rtx-4070-super` | Muramasa82 | NVIDIA GeForce RTX 4070 SUPER, driver NVIDIA 621.24, Windows 11 | 77 | | `sx-nn-rtx-4070-super` | sx.nn | NVIDIA GeForce RTX 4070 SUPER, driver NVIDIA 610.88, Windows 11 | 1 | Not added: MB44's list (292 pipelines) and piano_butcher's (142). Everything in them is already in other lists, as [#1688](https://git.magicbean.dev/MagicBeanDev/Longshot/pulls/1688) also found. How the zips were handled: - Each attachment was downloaded into a folder of its own, and its members were listed. - All nine held only `longshot_pipeline_states.bin`, and eight also held `longshot_pipeline_report.txt`. Only those members were read, and nothing was run. - Each list passed `nx2/pipeline_states.py --check` before `add`. All are version 2. - The notes' GPU, build and settings come from the zip's report, or from the comment for Arturito-604, whose zip has no report. The route comes from the comment, or "Not given" where the player left it blank. See #14 (it stays open for more lists). ## Testing **Tested:** `nx2/pipeline_states.py --check` passed on each of the nine sent lists. After the adds, `nx2/pipeline_states.py check` passes on the sources: `merged: 76567 pipelines (compute 3357, graphics 68877, mesh 4333) of 27276 programs, version 2`. That is under `COMPUTE_MOST` (3584). The test suite was not run on this PC. There was no game run: the PR was opened with `--force` at the maintainer's request, because it changes no game code, only the data the first start prepares. **Still to test:** - Any GPU, a fresh game folder built from this branch: let the first start's shader preparation run to the title, then play a few minutes. Expect it to finish without errors, and the log to report 555 more pipelines prepared than on main (by hand). ## Measurements **Verdict:** nothing measured in the game; the data only adds 555 pipelines the first start prepares. The counts below come from `nx2/pipeline_states.py check`. | | Before: main @ 7c23d2be | After: this PR @ d68cd654 | |---|---|---| | merged pipelines | 76,012 | 76,567 | | graphics | 68,393 | 68,877 | | mesh | 4,310 | 4,333 | | compute programs | 3,309 | 3,357 | | programs | 27,125 | 27,276 | ## Risks - **First start:** preparation takes a little longer, for 555 more pipelines (under 1 % more). - **Memory:** the 48 more compute programs are kept ready on every start. By the measurement in `nx2/pipeline_states.py`, that is about 1 MB and a few ms each, so roughly 50 MB and well under a second per start. - **Turning it off:** revert the commit. - **After updating:** a player's first start prepares the new pipelines once. Saves, options and caches are otherwise unchanged. ## Checklist ### Ready for review - [x] One topic, on current `main`; the title is a Conventional Commit. - [ ] Tested on my own PC, in the game for a change the game runs, as Testing says. (Not run in the game: data only, opened with `--force` at the maintainer's request, as #1688.) - [ ] `mise run test` passes (any failure that also fails on `main` is named). (Not run on this PC; only `nx2/pipeline_states.py check`, which passes.) - [x] Measured against `main`, or Measurements says why there is nothing to measure. - [x] No game files, build output or hand-edited generated files; `third_party/` untouched; nothing GPL-incompatible. ### When it applies - [x] AI assistance recorded with `Assisted-by:` / `Co-Authored-By:` trailers. ### Ready to merge - [ ] Everything in Still to test has run, its results in Tested. - [ ] Tier 1 hardware it can affect is tested (docs/dev/standards.md, "Hardware tiers"). - [ ] Approved in review, every comment answered. - [ ] CI passes on the head; on current `main`, measured again or a comment says why the numbers still hold. 🤖 Generated with [Claude Code](https://claude.com/claude-code)
build(release): 7 more contributed pipeline lists from #14, 555 pipelines
Some checks failed
CI / Python source checks (ubuntu-24.04, part 1/2) (pull_request) Has been skipped
CI / Python source checks (ubuntu-24.04, part 2/2) (pull_request) Has been skipped
CI / SPIR-V fixtures (global narrow load/store, direct surface atomics, F2I immediate) (pull_request) Has been skipped
Third-party parts / The same as upstream (pull_request) Has been skipped
CI / Conventional Commit title (pull_request) Successful in 11s
CI / Workflow lint (actionlint) (pull_request) Successful in 17s
Guide screenshots / Guide images and local UI source review (pull_request) Failing after 18s
Website / Website on Cloudflare (pull_request_target) Has been skipped
Third-party parts / As their notes say (pull_request) Successful in 21s
PR description / Description follows the template (pull_request_target) Failing after 5s
CI / Website build and checks (pull_request) Successful in 2m27s
d68cd654af
The lists players attached to the Submit your pipelines issue since #1688:
Arturito-604, bwen, icastot, jlh4426, Maver1ck, Muramasa82 and sx.nn. Each
passed nx2/pipeline_states.py --check and was added with add, its GPU,
build and settings from the zip's report and its route from the comment.
MB44's and piano_butcher's hold nothing the lists before them lack.

Assisted-by: Claude Code (Claude Opus 5.5)
Co-Authored-By: Claude Opus 5.5 <[email protected]>
HumanBean changed title from WIP: build(release): 7 more contributed pipeline lists from #14, 555 pipelines to build(release): 7 more contributed pipeline lists from #14, 555 pipelines 2026-11-21 07:53:05 +00:00
HumanBean deleted branch build/pipelines-from-14-2 2026-11-21 07:53:21 +00:00
Sign in to join this conversation.
No reviewers
No milestone
No project
No assignees
1 participant
Notifications
Due date
The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set

Reference
MagicBeanDev/Longshot!2041
No description provided.