Old promo URLs stay for broken links

A publisher that cleans up after itself still has to decide what happens to the addresses people already shared. Oernoe’s rule for that case sits in What we publish under pages that stay up but leave the index. Older promotional posts can remain at their URLs so existing links do not break. They get a noindex instruction, they drop out of the journal listing and the machine-readable index of public pages, and they do not load ads. That is not a quiet delete. It is a deliberate half-life: the document can still open if someone clicks an old bookmark, but it stops pretending to be part of the live journal.

This essay is only about that half-life. It is not a walkthrough of rewriting a thin SEO outline into an Oernoe-specific document. It is not a sermon about brochure copy on the homepage. It is not a checklist for spotting a publish slip. Other pieces already cover those jobs. Here the job is narrower: why a promotional URL is allowed to resolve after we have decided it should not rank, should not sit in the listing, and should not carry inventory.

## What “stay up” actually means

When we say a promotional URL stays up, we mean the server still returns a page at that address. The page is not a soft 404 dressed as a joke. It is also not a redirect maze that dumps the reader into Home and hopes they forget what they clicked. A person who typed the old address, or followed a link from a forum thread written months ago, should land on something that explains the situation. What we publish is explicit: if you open one of those URLs directly, you should see a note that it is kept for old links. You should not see it in the journal listing, the homepage “from the journal” block, or the public URL index.

That note is part of the honesty rule. A blank shell that still returns 200 is how reviewers decide a site is padded. A cheerful brochure that still sells a launch that never happened is how readers decide the company lies casually. The kept-for-old-links note sits between those failures. It admits the URL had a previous life. It refuses to promote that life as current. It also refuses to erase the address from the web as if history were optional.

The mechanical side matches the prose. noindex tells crawlers not to treat the page as a candidate for results. Removal from the journal listing tells humans the same thing without forcing them to read headers. Leaving ads off those pages tells anyone who audits placement that unfinished or superseded promotional material is not inventory. What we publish groups that treatment with test slugs, “coming soon” titles, and brochure posts we already identified. The pattern is the same whether the original mistake was a slogan about the future of free tech or a safest-platform claim that could not survive contact with how this publisher actually funds selected pages.

## Why breaking links is worse than leaving a noindex page

It is tempting, when a team is embarrassed by old copy, to delete the file and congratulate itself on hygiene. Link rot is not hygiene. It is a tax on every third party that trusted the URL when it was current: a citation in a support reply, a bookmark in a browser, a mention in a thread that still ranks elsewhere. Those people did not agree to a scavenger hunt. They agreed to an address.

Keeping the address with a clear note is the cheaper ethical choice for a small publisher. It costs a few lines of template and a policy sentence. Rewriting every inbound reference on the open web is not available to us. Returning a hard 404 on a URL that once carried a public claim also trains people to assume Oernoe disappears pages when they become inconvenient. We already publish Corrections for claims we took down or rewrote. The promotional half-life is the cousin of that list: not every old URL needs a corrections entry, but every old URL that still resolves needs a status that matches reality.

Leaving a promotional page in the index while stripping its meaning creates a worse object than a noindexed kept page: an indexed brochure with nothing to say. Taking it out of the index while leaving the URL for humans who already have the link matches both the quality rule and the link-stability rule.

## What kinds of pages get this treatment

What we publish is concrete about the bucket. Older promotional posts. Test slugs. Coming-soon titles. Brochure posts we already identified. The same treatment applies to posts written as if a stranger can open products that are not live on the homepage today. We are not pretending those URLs never existed. We are refusing to offer them as the site.

That last clause matters for how we write about unfinished work. The homepage directory is the authority for what is live. About says the same thing in different clothes: if a page here disagrees with the homepage directory, the homepage is the one to trust until we fix the page. Journal posts that described unreleased application surfaces as if they were open products were not aspirational marketing. They were false product claims with durable URLs. The recovery is not to keep selling the claim in softer language. The recovery is to leave the URL for old links, mark it noindex, strip it from listings, keep ads off, and point readers toward what is actually available.

Brochure language fails for a related reason. A page whose only job is to sound like a launch announcement does not become a journal article by growing filler. Word count is a floor on What we publish, not a rescue for a brochure. When a brochure URL cannot be rewritten into something that earns the index, the promotional half-life is the remaining honest option. Test and coming-soon pages fail even faster: leaving them in the public listing teaches reviewers that the journal is a staging area.

## Distinct from neighboring recoveries

Rewriting a path into an Oernoe-specific document is a different recovery. Four old journal URLs used to be generic SEO outlines. Those URLs still exist, and the documents on them now earn their keep about what we crawl and how we publish. That is a rewrite path. The promotional half-life is for pages that should not return to the index as they stand.

Brochure pages that do not clear the floor are a quality judgment about readiness. This essay assumes that judgment has already gone against the page. The question here is disposal of the URL after that judgment, not the judgment itself.

How to tell if we slipped on publish is an operational sniff test: teaser walls, generic SEO returns, purity claims on Home that disagree with ads on finished guides. Those checks tell you whether the live site has drifted. They do not explain why a noindexed promotional URL is still allowed to open.

Robots headers on application shells that should not rank are another cousin. Shells that are not documents send noindex so a crawler that can fetch the URL still knows not to treat the response as inventory. Promotional journal URLs use the same family of exclusion, but the human-facing kept-for-old-links note and the drop from the journal listing are specific to cleanup of old promo writing. Promotional exclusions as a quality tool is the policy umbrella; this essay is one worked example under it.

## What a reader should see

If you follow an old promotional link today, the honest experience is boring on purpose. You get a page. You get a note that the URL is kept for old links. You do not get a soft sell into a waitlist that does not exist. You do not get an ad unit glued to a slogan. You do not get the article promoted in the journal rail as if it were current editorial. You do not get the URL offered in the public machine-readable list of pages we want discovered.

If you arrive through the journal listing or the homepage block that surfaces finished articles, you should not meet those promotional URLs at all. Absence from those surfaces is part of the product. Presence at the raw address is a compatibility layer for the open web’s memory, not an invitation to treat the old brochure as the company.

That split also protects the ads story we tell elsewhere. Search queries and Google ads and the Privacy Policy both draw a hard line between selected finished publisher pages that may load AdSense and pages that must not. Thin stubs, login, legal, and the journal listing stay off inventory. Promotional leftovers belong with that exclusion set. Monetizing a kept-for-old-links page would turn a cleanup into a second mistake.

## Recovery rule, refusals, and the editor check

What we publish opens by saying www.oernoe.com is allowed to be a publisher of original writing about search, privacy, and the products we actually run; a set of help pages; and a legal and company record. It is not allowed to be a pile of coming-soon notes, generic SEO essays, or slogans about being the safest platform of the year. The promotional half-life enforces that refusal without pretending the earlier URLs were never public.

Corrections names related failures: homepage purity language that could not survive AdSense on selected guides, interest capture for software that was not released, generic SEO outlines, robots rules that did not match live routes. Not every excluded promo needs a corrections paragraph. Every excluded promo needs the same mechanical honesty: noindex, out of the listing, no ads, and a note if the page still opens. About reinforces the authority chain: product claims check against the homepage directory; policy claims check against Privacy, Cookies, and Terms.

We will not quietly redirect every old promo into Home and call that a fix. We will not leave the old slogan without the kept-for-old-links note. We will not put those URLs back into the journal listing because nostalgia traffic looks useful. We will not load ads on them to make the cleanup pay for itself. We will not claim that leaving a URL up means the product described there is live. The homepage directory decides liveness. The old URL decides only whether a prior address still answers.

Before you treat a promotional URL as finished business, ask four questions. Does the raw address still need to answer for inbound links? If it must, write the kept-for-old-links note in plain language. Is noindex present, and is the URL absent from the journal listing and the public URL index? Are ads off? Does the page still speak as if an unreleased product is open today? If it does, the note is not enough; the body has to stop making the false claim even while the URL remains for old links.

Those checks are dull. Dull is the point. The glamorous version of cleanup is a rewrite that earns the index again. The necessary version, for pages that will not earn it, is a stable address with no ranking ambition and no inventory. What we publish already chose that necessary version for older promotional posts. Old promo URLs stay for broken links. They do not stay for discovery, listing, or ads. That is the whole rule, and it is enough.