Repository navigation
releases: the search text calls 1bit OS an appliance, not 'the live image' - #32
Conversation
…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
left a comment
There was a problem hiding this comment.
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.
|
VERDICT: READY One word, and the PR quotes the promise it is keeping — "site#16's body lists Dropping 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 |
|
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. 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: 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 |
The one sentence site#16 promised and no PR contains
site#16's body lists
tools/site.py:The CHANGES on it is right that
tools/site.pyis 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: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:One line
158 → 143 characters, because the page's own check enforces the limit:
Verified:
ONEBIT_ENGINE_DIR=<engine checkout> python3 tools/site.py _site→ 72 pages, rc 0, and the description in_site/releases.htmlno 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).