Skip to content

docs: link the published device driver catalog - #720

Merged
frahlg merged 1 commit into
srcfl:masterfrom
HuggeK:link-driver-catalog
Aug 4, 2026
Merged

docs: link the published device driver catalog#720
frahlg merged 1 commit into
srcfl:masterfrom
HuggeK:link-driver-catalog

Conversation

@HuggeK

@HuggeK HuggeK commented Jul 30, 2026

Copy link
Copy Markdown
Contributor

The README tells a reader to send driver changes to srcfl/device-drivers, but never says which devices are already covered. That repository now publishes a searchable catalog at https://srcfl.github.io/device-drivers/, generated from its manifests and driver sources on every push to main.

This links it from two places:

  • the Drivers section, next to the existing pointer at srcfl/device-drivers;
  • the Documentation index.

The page shows all 80 drivers with their protocol, emitted DER types, control capability, tier and version, plus every tested model family — searchable by manufacturer or model number, so SH10RT finds sungrow.

It states hardware evidence rather than implying it: 5 drivers report being confirmed against physical hardware and 28 report in their own verification_notes that they have not been. Those are the same drivers and versions FTW installs from the signed channel.

Note

The catalog is live. srcfl/device-drivers#51 merged and its Pages workflow has run — https://srcfl.github.io/device-drivers/ serves 200, so the link does not 404 and this is ready to merge.

Counts verified against the published page

The three figures above were re-checked against the live catalog rather than restated from the source PR: 80 drivers, 5 with hardware_verified: true, and 28 whose verification block reports level: unverified / "Not verified on hardware".

The README itself deliberately states no numbers — only that the page distinguishes confirmed from unconfirmed — so it cannot go stale as drivers are added.

Why the changeset check failed earlier

Not a problem with this PR. The old changeset-check workflow ran changeset status before deciding the docs allowlist, and changeset status reports "packages have been changed but no changesets were found" for any diff inside a package — a README-only diff included.

#721 reordered the gate so the allowlist is decided first. That merged 19 minutes after this PR's last run, so the branch has been updated onto it.

Why no Changeset

Documentation only — no behaviour, API, config or UI change, so there is nothing for a release note to describe.

🤖 Generated with Claude Code

@frahlg

frahlg commented Jul 30, 2026

Copy link
Copy Markdown
Member

Hold this one — the link target does not resolve yet.

$ curl -sIL https://srcfl.github.io/device-drivers/
HTTP 404

$ gh api repos/srcfl/device-drivers/pages
{"message":"Not Found","status":"404"}

GitHub Pages is not enabled on srcfl/device-drivers, so the catalog has no published site behind it. Merging this would put a dead link in the README's Drivers section and its Documentation index.

That matters more than usual today. The site's four install buttons and its closing CTA had been pointing at five README anchors deleted months ago — every visitor who clicked one landed at the top of the README with no instructions in view, and nothing noticed. srcfl/ftw-web#6 fixed it this morning; #719 is building the guard so it cannot recur. Adding a fresh dead link on the same day would be a poor trade.

The same dependency blocks srcfl/ftw-web#7, which points "Browse drivers" at the same URL, and srcfl/device-drivers#51 is still an open draft.

Order that works: enable Pages on srcfl/device-drivers and land #51, confirm the URL returns 200, then this and ftw-web#7 together. The content of this PR reads fine — it is only the timing.

@HuggeK
HuggeK marked this pull request as ready for review July 30, 2026 10:41
@HuggeK

HuggeK commented Jul 30, 2026

Copy link
Copy Markdown
Contributor Author

Done and merged in the device driver repo.

@frahlg

frahlg commented Jul 30, 2026

Copy link
Copy Markdown
Member

Retracting the hold — the blocker is gone.

$ curl -sIL https://srcfl.github.io/device-drivers/
HTTP 200

Pages is live on srcfl/device-drivers since I posted that comment, so the catalog this PR links to now resolves. My earlier objection was the only thing I had against it; the content itself I said read fine, and that still holds.

Same applies to srcfl/ftw-web#7, which points "Browse drivers" at the same URL and was waiting on the same dependency. Note those two conflict with each other on the "Write a driver" / "Browse drivers" block — ftw-web#6 already landed and rewrote it, so #7 needs a rebase before it will merge.

This one is BEHIND master (v1.14.0 shipped in the meantime), so it needs an update before merge, but nothing here blocks it.

@miravoss26 miravoss26 left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Adds a README paragraph and a docs-list entry linking the published device catalog (srcfl.github.io/device-drivers), with the useful caveat that a listing isn't an install claim.

  • Docs-only, one file. CI is green (go test + vet passed, the rest path-skipped).
  • No secrets, deps, or network/authz surface. The link is the org's own Pages domain.

Accurate and consistent with the catalog wording elsewhere. Safe to merge from my read. (mergeable shows UNKNOWN, but that's GitHub not having recomputed yet; nothing in the diff blocks it.)

frahlg added a commit that referenced this pull request Aug 4, 2026
…uide

FTW and srcfl/device-drivers both carried a "how to write a driver" guide,
220 and 268 lines. They overlap on the entry points, the DRIVER block, the
sign convention and the build-and-test loop, and there is no mechanism that
keeps them agreeing. The upstream copy is built around blueprint/BLUEPRINT.lua
and is the one a driver author reaches for, so the FTW copy is the one that
quietly goes stale — and #720 is currently moving the website's "Write a
driver" link upstream, which makes that worse.

Split by what each repository actually owns. Authoring moves upstream, and
this page keeps what only FTW can answer: what its host registers, how a
capability is granted, when the host calls driver_default_mode, where FTW
resolves a driver file from, and how to test one against a running instance.

The host API is now a table generated from what lua.go registers, including
http_patch and its allow_write gate from #716. Every name in it was checked
against host.RawSetString in lua.go; the two Blixt L1 aliases stay in the
prose below the table rather than the table itself.

Removed content is covered elsewhere: entry points, sign convention, manifest
and make targets upstream, and the drivers-beta promotion paragraph was
already a duplicate of device-repository.md, which this page links to.

README is untouched — #720 is editing the same section and has right of way.

Co-authored-by: Claude Opus 5 <noreply@anthropic.com>
The README told a reader to send driver changes to srcfl/device-drivers
but never said which devices are already covered. That repository now
publishes a searchable catalog, generated from its manifests and driver
sources on every push, so this points at it from the Drivers section and
the documentation index.

The page states hardware evidence rather than implying it: 5 of the 80
drivers report being confirmed against physical hardware and 28 report
that they have not been, which matches what FTW actually installs.

Documentation only, no user-visible change, so no Changeset.

Signed-off-by: Hugo Karlsson <48095810+HuggeK@users.noreply.github.com>
Co-authored-by: HuggeK <48095810+HuggeK@users.noreply.github.com>
@frahlg
frahlg force-pushed the link-driver-catalog branch from 9c69092 to f5a2fae Compare August 4, 2026 07:38
@frahlg
frahlg merged commit 3f8baf3 into srcfl:master Aug 4, 2026
13 checks passed
frahlg added a commit that referenced this pull request Aug 4, 2026
…uide

FTW and srcfl/device-drivers both carried a "how to write a driver" guide,
220 and 268 lines. They overlap on the entry points, the DRIVER block, the
sign convention and the build-and-test loop, and there is no mechanism that
keeps them agreeing. The upstream copy is built around blueprint/BLUEPRINT.lua
and is the one a driver author reaches for, so the FTW copy is the one that
quietly goes stale — and #720 is currently moving the website's "Write a
driver" link upstream, which makes that worse.

Split by what each repository actually owns. Authoring moves upstream, and
this page keeps what only FTW can answer: what its host registers, how a
capability is granted, when the host calls driver_default_mode, where FTW
resolves a driver file from, and how to test one against a running instance.

The host API is now a table generated from what lua.go registers, including
http_patch and its allow_write gate from #716. Every name in it was checked
against host.RawSetString in lua.go; the two Blixt L1 aliases stay in the
prose below the table rather than the table itself.

Removed content is covered elsewhere: entry points, sign convention, manifest
and make targets upstream, and the drivers-beta promotion paragraph was
already a duplicate of device-repository.md, which this page links to.

README is untouched — #720 is editing the same section and has right of way.

Co-authored-by: Claude Opus 5 <noreply@anthropic.com>
frahlg added a commit that referenced this pull request Aug 4, 2026
…uide

FTW and srcfl/device-drivers both carried a "how to write a driver" guide,
220 and 268 lines. They overlap on the entry points, the DRIVER block, the
sign convention and the build-and-test loop, and there is no mechanism that
keeps them agreeing. The upstream copy is built around blueprint/BLUEPRINT.lua
and is the one a driver author reaches for, so the FTW copy is the one that
quietly goes stale — and #720 is currently moving the website's "Write a
driver" link upstream, which makes that worse.

Split by what each repository actually owns. Authoring moves upstream, and
this page keeps what only FTW can answer: what its host registers, how a
capability is granted, when the host calls driver_default_mode, where FTW
resolves a driver file from, and how to test one against a running instance.

The host API is now a table generated from what lua.go registers, including
http_patch and its allow_write gate from #716. Every name in it was checked
against host.RawSetString in lua.go; the two Blixt L1 aliases stay in the
prose below the table rather than the table itself.

Removed content is covered elsewhere: entry points, sign convention, manifest
and make targets upstream, and the drivers-beta promotion paragraph was
already a duplicate of device-repository.md, which this page links to.

README is untouched — #720 is editing the same section and has right of way.

Co-authored-by: Claude Opus 5 <noreply@anthropic.com>
frahlg added a commit that referenced this pull request Aug 4, 2026
…uide

FTW and srcfl/device-drivers both carried a "how to write a driver" guide,
220 and 268 lines. They overlap on the entry points, the DRIVER block, the
sign convention and the build-and-test loop, and there is no mechanism that
keeps them agreeing. The upstream copy is built around blueprint/BLUEPRINT.lua
and is the one a driver author reaches for, so the FTW copy is the one that
quietly goes stale — and #720 is currently moving the website's "Write a
driver" link upstream, which makes that worse.

Split by what each repository actually owns. Authoring moves upstream, and
this page keeps what only FTW can answer: what its host registers, how a
capability is granted, when the host calls driver_default_mode, where FTW
resolves a driver file from, and how to test one against a running instance.

The host API is now a table generated from what lua.go registers, including
http_patch and its allow_write gate from #716. Every name in it was checked
against host.RawSetString in lua.go; the two Blixt L1 aliases stay in the
prose below the table rather than the table itself.

Removed content is covered elsewhere: entry points, sign convention, manifest
and make targets upstream, and the drivers-beta promotion paragraph was
already a duplicate of device-repository.md, which this page links to.

README is untouched — #720 is editing the same section and has right of way.

Co-authored-by: Claude Opus 5 <noreply@anthropic.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

4 participants