Conversation
The remaining two instances of #3355's class, both repaired in the shape #3356 introduced: the refusal goes into the file the tool already reads. `tri fpga program-flash` advertised `--enable-quad` as "needed for some boards/flash", with `--part` defaulting to `xc7a200tfgg676` -- the exact board `fpga/HARDWARE_SSOT.md:323` refuses both quad flags for, as a MEASURED result (W396, experiment E4): the Micron N25Q128_3V has no separate QE bit, so openFPGALoader aborts. Measured: zero guards, warnings or mentions of the part anywhere near the flag; the pass-through was bare. The runtime line was worse than the flag. All three internal callers hardcode `--spi-buswidth 1`, the SSOT-canonical no-quad path, and on it the command printed "ensure the flash QE bit and board straps match" -- sending a reader to align a bit the SSOT records as not existing on this part, in the middle of the experiment the SSOT itself prescribes. Silent on x1 now. `quad_refused()` is a predicate rather than an inline condition SO THAT it can be tested without reaching the code that spawns `openFPGALoader`: a test that could spawn the programmer is a test that writes to a board when it regresses. `tri gates dead` printed, for every suppressed row, that "a workflow whose last step is forbidden by this repository's own ruleset fails every time it runs". On master `brain-seal-refresh.yml`'s last step is `Upload brain seals` and the only `git push` left in the file is inside the comment explaining its removal by #3324. The footnote had no surviving referent, and three further assertions in `gates.rs` still stated the old cause; all four corrected. `# tri:cause-removed <reason>` gives the classifier the fourth state the domain has -- never green, cause diagnosed and removed, no run since. The row is listed with its reason, and the blanket footnote is scoped to the rows that explain nothing: 3 of 4 today. A bare marker with no reason is not a reason. Two of my own traps on the way, both already in my notes and hit again: an insertion anchored on `pub enum Bucket` landed between the enum and its `#[derive]`, and two edits keyed on the wrong indentation reported success while writing nothing. 814 crate tests pass; 0 fmt diffs in the two files touched, and `cargo fmt` was NOT run, because it reformats 67 pre-existing sites elsewhere. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Contributor
Contributor
|
📓 NotebookLM Notebook linked to this PR
This notebook contains session context, decisions, and artifacts for this work. |
Contributor
PR DashboardGenerated at: 2026-09-06 07:16:54 UTC
Summary
Seal Status
|
Contributor
|
📓 NotebookLM Notebook linked to this PR
This notebook contains session context, decisions, and artifacts for this work. |
Contributor
|
📓 NotebookLM Notebook linked to this PR
This notebook contains session context, decisions, and artifacts for this work. |
Contributor
PR DashboardGenerated at: 2026-09-06 07:39:13 UTC
Summary
Seal Status
|
gHashTag
enabled auto-merge
September 6, 2026 08:02
Contributor
PR DashboardGenerated at: 2026-09-06 08:03:44 UTC
Summary
Seal Status
|
Contributor
|
📓 NotebookLM Notebook linked to this PR
This notebook contains session context, decisions, and artifacts for this work. |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Refs #3371
The remaining two instances of #3355's class, both repaired in the shape #3356 introduced: the refusal goes into the file the tool already reads.
1.
tri fpga program-flashadvertised a flag its own bench refuses--enable-quadsaid "Needed for some boards/flash to boot from x4 SPI", with--partdefaulting toxc7a200tfgg676— the exact boardfpga/HARDWARE_SSOT.md:323refuses both quad flags for, as a measured result (W396, experiment E4): the Micron N25Q128_3V has no separate QE bit, so openFPGALoader aborts. Measured: zero guards, warnings or mentions of the part anywhere near the flag.The runtime line was worse than the flag. All three internal callers hardcode
--spi-buswidth 1, the SSOT-canonical no-quad path, and on it the command printed "ensure the flash QE bit and board straps match" — sending a reader to align a bit the SSOT records as not existing on this part, in the middle of the experiment the SSOT itself prescribes.quad_refused()is a predicate rather than an inline condition, so it can be tested without reaching the code that spawnsopenFPGALoader. A test that could spawn the programmer is a test that writes to a board when it regresses.2.
tri gates deaddiagnosed a cause a merged PR removedIt printed, for every suppressed row, that "a workflow whose last step is forbidden by this repository's own ruleset fails every time it runs". On master
brain-seal-refresh.yml's last step isUpload brain seals, and the onlygit pushleft in the file is inside the comment explaining its removal by #3324. The footnote had no surviving referent; three further assertions ingates.rsstill stated the old cause. All four corrected.# tri:cause-removed <reason>gives the classifier the fourth state the domain has. Live now:A bare marker with no reason is not a reason — asserted, because it would print "cause removed:" and explain nothing.
Mine, on the way
Two traps already in my notes and hit again: an insertion anchored on
pub enum Bucketlanded between the enum and its#[derive], and two edits keyed on the wrong indentation reported success while writing nothing.814 crate tests pass. 0 fmt diffs in the two files touched — and
cargo fmtwas deliberately not run, because it reformats 67 pre-existing sites elsewhere.