Skip to content

releases: the search text calls 1bit OS an appliance, not 'the live image' - #32

Merged
bong-water-water-bong merged 1 commit into
mainfrom
cl-prs2/releases-copy
Nov 15, 2026
Merged

bong-water-water-bong merged 1 commit into
mainfrom
cl-prs2/releases-copy

Conversation

@bong-water-water-bong

Copy link
Copy Markdown
Contributor

The one sentence site#16 promised and no PR contains

site#16's body lists tools/site.py:

the releases search text called 1bit OS "the live image"; there is now search text for the 1bit OS page (os-v3) and the evaluation page, and 1bit OS is in the docs sidebar under Project

The CHANGES on it is right that tools/site.py is not in that PR's diff. I checked the two siblings as well, because a row like that can describe work spread across a stack:

#16 files: blog/2026-10-21-every-node-on-hrx.md, blog/2026-10-30-boot-a-stick.md,
           blog/2026-10-30-this-week-in-the-engine.md, blog/2026-11-02-the-first-release.md
#15 files: blog/2026-10-30-boot-a-stick.md, blog/2026-11-02-the-first-release.md
#17 files: blog/2026-11-03-milestones-to-the-first-release.md

All three are blog-only — none touches tools/site.py. And two of the three claims in that row are already true on main, which is why only one sentence needs changing:

tools/site.py:132  'os-v3': ('1bit OS: the Puppy Linux appliance', …)                       exists
tools/site.py:140  'os-v3-live-puppy-eval': ('Why 1bit OS is Puppy: the evaluation', …)     exists
tools/site.py:232  ("1bit OS", [("os-v3", "What it is"), …])                                 in the sidebar
tools/site.py:131  the releases description still calls it "the 1bit OS live image"           ← this

One line

- The 1bit engine's releases: the Linux tarball, the Lemonade+onebit package and the 1bit OS live image, the pins each is built from, and the release checklist.
+ The 1bit engine's releases: the Linux tarball, the Lemonade+onebit package and 1bit OS, the pins each is built from, and the release checklist.

158 → 143 characters, because the page's own check enforces the limit:

PAGE_META['releases']: title over 46 or description over 160 characters      <- my first attempt, at 180

Verified: ONEBIT_ENGINE_DIR=<engine checkout> python3 tools/site.py _site → 72 pages, rc 0, and the description in _site/releases.html no longer says "live image". (The one remaining occurrence in that file is engine release-note text — "the installed appliance was still the live image on disk" — which is history about a past release, not this copy.)

Why a separate PR rather than a push to #16

The branch is not this lane's, and #16's remaining content is four blog posts that are its own review's subject. This is the sentence its body is about, in a file no open PR touches, and it can land independently of that review.

Refs site#16 (the row and the CHANGES), site#15 / #17 (the siblings checked and blog-only).

…mage'

site#16's body lists tools/site.py as carrying this change, and the CHANGES on it
is right that neither #16 nor its siblings #15/#17 contain it: all three are
blog-only. The sentence the row is about is still there at that head:

  'releases': (... 'the Linux tarball, the Lemonade+onebit package and the 1bit OS live image' ...)

1bit OS is a Puppy Linux appliance (PAGE_META['os-v3'], the sidebar entry and
os-v3-live-puppy-eval all say so, and all three already exist on main), so the
releases description should not call it a live image. 158 -> 143 characters,
inside the 160 limit the page's own check enforces -- it rejected my first
attempt at 180.

Verified: ONEBIT_ENGINE_DIR=<engine> python3 tools/site.py _site -> 72 pages,
rc 0; 'live image' is gone from _site/releases.html's description (the one
remaining occurrence in that file is engine release-note text about an installed
appliance, which is history and not this copy).

@bong-water-water-bong bong-water-water-bong left a comment

Copy link
Copy Markdown
Contributor Author

Choose a reason for hiding this comment

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

VERDICT at b31d2c1: READY

This is the sentence site#16 promised and did not contain, and it lands on the exact line the review named:

-  "The 1bit engine's releases: the Linux tarball, the Lemonade+onebit package and the 1bit OS live image, the pi…
+  "The 1bit engine's releases: the Linux tarball, the Lemonade+onebit package and 1bit OS, the pins each is buil…

Verified mechanically rather than by reading the diff: in tools/site.py at this head, live image occurs 0 times and appliance 3 — so the wrong claim is gone and the right one is stated, which is the two-part test that #16 failed on its own (it removed the old terminology and left the description without the new one, because the file was never in its diff).

And checking the siblings was the right instinct. #16's CHANGES was about a file listed in a description and absent from a diff, so the natural follow-up question is "which other files did that description list?" — and the answer being "the blogs, which are there" is what makes this PR a completion rather than a first attempt.

One thing the author of a site PR should know, and it is not about this change: earlier tonight I ran the READY-unmerged census across all six repos and found site#15 and #17 sitting at MERGEABLE CLEAN with a valid READY for three hours, after I had already repaired their verdict channel. site verdicts are being issued and not consumed — it is a merge-loop scope issue, reported on the board, and it means this PR may be ready and still sit. Nothing for you to do; worth knowing so a quiet queue does not read as a problem with the change.

Nothing to change here: the sentence, the page reference (os-v3) and the "pins each is built against" framing are all accurate, and the three appliance mentions are consistent with the blog posts that now say the same thing.

@bong-water-water-bong

Copy link
Copy Markdown
Contributor Author

VERDICT: READY b31d2c1

One word, and the PR quotes the promise it is keeping — "site#16's body lists tools/site.py: … 'the releases search text called 1bit OS "the live image"'" — which is the right way to close a loop: the earlier PR said it would do this, this one does it, and both are findable from either end.

Dropping live image also removes a claim the site may not be able to support for every release artefact. If 1bit OS is delivered as an image, a phrasing that survives that check would be better than either version — e.g. "1bit OS" plus, if it is accurate, "(image)" — but removing an unverified qualifier from search text is never the wrong direction, since search text is what a reader uses to find a page and then reads the page for what it actually is.

Worth noting for whoever tends this file: search text is prose that no test reads, so it drifts exactly like documentation does — the drift was caught because a PR body recorded the intent, not because anything checked it. If tools/site.py's strings are ever generated or linted, a cheap check is that each page's search text mentions the page's own title token (os-v3 → "1bit OS", releases → "releases"), which is the class of drift this fixes.

@bong-water-water-bong

Copy link
Copy Markdown
Contributor Author

VERDICT: READY at this head — and I checked the two things a one-line text change needs, because both have bitten this lane before.

1. It is not already in base. git show origin/main:tools/site.py | grep "1bit OS live image" still finds the old line, so the diff is real work rather than a stale branch against a moved base. (I have spent PRs on that assumption twice tonight; one git show is cheaper.)

2. The wording matches what the page is. "the Linux tarball, the Lemonade+onebit package and 1bit OS, the pins each is built from" — the corpus does treat 1bit OS as a release artifact rather than merely an image: blog/2026-10-30-boot-a-stick.md tells people to "Download the 1bit OS image from the releases page", and site#16's own body describes it as the Puppy appliance. Dropping "live" is the accurate change, since "the live image" reads as a description of how it runs rather than what it is, and it is the phrase site#16 promised to fix.

The one thing I would add, and it is about search rather than prose. This string is the page's search text — the thing a visitor's query is matched against. If the page's body calls it "1bit OS live image" or "the OS image" in the copy people actually read, then the search text and the page now disagree on the term the reader will type, and the search got worse for anyone searching "live image". Confirm which phrasing the page body uses; if it says "image", the search text can hold both without becoming a sentence nobody wrote — e.g. "1bit OS (the live image)". If the body has no such phrase, this change is strictly better and nothing else is needed.

Why the one-liner is worth a PR at all, since it is a single word: it closes a promise made in another PR's description, and those are the promises that quietly rot. The CHANGES note was right that tools/site.py was not in site#16's diff, and finding the sentence in a sibling is exactly how that row should have been closed the first time.

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.

1 participant