Privacy-first crawl is not infinite index

How Oernoe Search works puts four crawl bullets in one list: continuous crawling, respecting robots.txt and crawl rate limits, privacy-first crawling that does not store identifying data during the fetch, and content analysis of pages and links. Readers sometimes hear the third bullet as if it enlarged the other three. It does not. Privacy-first crawl is a claim about what identifying data is not stored while a page is fetched. It is not a claim that the index holds every public URL. Search’s tip footer already says results come from websites and knowledge entries in the Oernoe index, and that not every page on the web is indexed. This essay is Search limits writing about that cut. It is not a remake of the rate-limits-as-politeness essay, not a remake of the finite-crawl-is-not-coverage-promise essay, and not a remake of the finite-index-means-missing-is-ordinary essay. Those hinges already exist. The hinge here is narrower: privacy language about the crawl does not mint an infinite store.

## What the guide actually claims about privacy-first crawling

The crawl section is careful. Crawlers visit pages they already know about, follow public links, and come back when a page changes. The index is what lets Search return results without asking every site on the web in real time. Continuous work keeps the store current for pages inside the discovery graph. Robots and rate limits keep that work polite. Content analysis looks at page content, structure, and links.

Privacy-first crawling, in the same list, refuses a different temptation. The guide says Oernoe does not store IP addresses, device information, or any identifying data during crawling. It says the company does not harvest email addresses for advertising lists, track user identities, or store behavioral signals as part of that fetch. It indexes publicly visible content and how pages link to each other. Those sentences are about the crawler’s data diet. They are not about how large the resulting index becomes.

The ranking section continues the same discipline from another angle. Content relevance, authority, link analysis, freshness, aggregated satisfaction signals, mobile-friendliness, and page speed can order documents the index already holds. Notice what is not on the list: your search history, your location, your device type, your age, your interests, or other personal advertising-profile inputs. Privacy-preserving ranking is a claim about how stored documents are ordered and about what personal advertising files are not built. Ranking prose cannot rank a page the crawler never fetched. Privacy-preserving ranking therefore cannot invent coverage either.

## The infinite-index fantasy that privacy language invites

People translate “privacy-first” into “careful,” then translate “careful” into “complete.” The leap feels moral. If the crawler is not building advertising lists, maybe it is free to visit everything. If Search does not profile queries, maybe absence is a temporary embarrassment rather than a design limit. That leap is how privacy marketing becomes a coverage promise the machine never made.

An infinite index would require visiting every public URL the moment it appears, forever, under every robots rule, at every language and mirror, with perfect extraction, and with no quality exclusions. No working search engine ships that fantasy. How Search works already refuses size-match marketing. Oernoe Search is a working search engine at search.oernoe.com. It is not a claim that the index matches every other index in size. The difference the guide will stand behind is narrower: Search does not use personal advertising profiles to rank results, and the publisher site discloses Google ads when they appear on selected pages. Size humility and privacy discipline travel together. A finite index that refuses query-based ad profiling is still finite.

Privacy-first crawling that refused to store identifying data during the fetch can still skip a host. It can still honour a Disallow line. It can still lag on a brand-new page with no inbound links. It can still leave spam and empty shells out. None of those absences are cured by the fact that the fetch did not build an advertising list.

## Distinct from three nearby essays

The rate-limits essay taught that crawl rate limits and robots respect are features, not outages. Its hinge was politeness. This page assumes politeness and asks a different question: whether privacy-first wording about the fetch enlarges the store.

The finite-crawl essay taught that crawl description is not a promise of coverage. Its hinge was honesty about the crawl process versus coverage marketing. This page zooms to one bullet inside that process — the privacy-first claim — and refuses the specific mistranslation that turns data-diet language into universal indexing.

The finite-index essay taught that missing is ordinary for any real stored set. Its hinge was acceptance of absence. This page explains why privacy-first crawl language is one of the stories that most often tempts people to reject that acceptance. Ordinary missing remains true even when the crawler’s privacy story is also true.

Three essays can share a guide without sharing a job. Confusing them produces thin journal volume: length without a new reader question. Corrections already named overlapping privacy essays as that failure mode on the publisher side. Search-limits writing should not repeat the pattern.

## Crawl privacy is not query privacy, and neither is coverage

How Search works separates several privacy claims that reviewers like to mash.

Query privacy: Search does not use queries to build advertising profiles. Queries are not sold as an advertising file. That claim lives on Search and in Privacy’s no-sale language.

Publisher advertising: selected finished pages on www.oernoe.com may show Google ads. Google may process page context, cookies or similar identifiers, IP address, browser and device information, and ad interaction data on those pages. Ads are requested as non-personalized by default. The cookie bar is a notice, not consent. That claim lives on the publisher hostname and in Privacy and Cookies.

Crawl privacy: during fetch, identifying data is not stored as part of building advertising lists from the crawl. Public content and links are what enter the index.

Coverage: the index is a subset of the web. Not every page is indexed.

Mashing those four into one slogan is how “privacy-first” becomes false comfort. A reader can have query privacy and still see an empty hit list. A site owner can enjoy polite crawl and still be absent. A publisher page can load AdSense while Search still refuses to turn queries into ad profiles. A privacy-first crawl bullet can be true while the tip footer’s finite-index sentence stays true. The useful habit is to keep the claims in separate boxes.

## Funding does not buy infinity either

The guide’s funding answer matters here. Oernoe uses optional premium features and contextual Google ads on selected publisher pages such as that guide. Search queries are not sold or used to build advertising profiles. Ads on the guide are disclosed in Privacy and Cookies. That model funds a publisher and labeled product tiers. It does not fund infinite aggressive crawl, and it does not convert privacy-first crawl language into a guarantee that every public page must appear.

Privacy Policy section 1 scopes advertising disclosures to selected publisher pages on www. They do not mean advertising is enabled on Chat, Drive, Docs, Health, Tracker, Account, login screens, or other application surfaces. About repeats the split: ads do not appear on the product hosts, Home, About, legal pages, or the journal listing. Expecting publisher ads on www to force universal indexing is a category error in the same family as expecting ads.txt to force a unit on Contact. Expecting privacy-first crawl wording to force the same universal indexing is the twin error with prettier branding.

## What privacy-first crawl does buy

It buys a checkable refusal. The crawler’s job is public content and links, not an advertising harvest of addresses or identities during the fetch. It sits next to robots respect so a privacy story cannot be used as an excuse to ignore publisher rules. It sits next to ranking-without-profiling so the company is not allowed to say “we are private” while building a search-history advertising file.

Those refusals are valuable. They are also finite. A refusal is not a mirror of the web. A refusal is a boundary around what the system is allowed to keep while it builds a bounded index.

## Everyday misreadings to refuse

“I cannot find my new page, so privacy-first crawl failed.” No. Discovery lag is ordinary. Privacy-first crawl did not promise same-hour inclusion.

“My robots file blocks crawlers, but private search should still show me.” No. Honouring robots is part of the same guide that contains the privacy-first bullet. Privacy does not override Disallow.

“Privacy-first means you crawl more carefully, so you must crawl more.” No. Careful about identifiers is not the same as larger about coverage.

“Search does not profile me, therefore every public company must have a Knowledge row.” No. Knowledge and Websites are different stores. Empty in one tab is not proof the privacy story collapsed.

“The publisher corrected zero-tracking slogans, so crawl coverage must now be total.” No. Corrections fixed a network-wide tracking overclaim versus AdSense disclosures. That correction did not enlarge Search’s index. It made publisher policy honest.


## Publisher index discipline is not Search coverage either

www.oernoe.com keeps a public journal index that prefers original, finished documents and removes or noindexes thin promotional leftovers. search.oernoe.com keeps a web and Knowledge index that is larger than the journal and still finite. Both disciplines produce missing pages on purpose. Confusing them produces bad bug reports: people ask support@ why a noindexed promotional journal slug does not rank in Search as if Search were obligated to resurrect what the publisher deliberately retired, then wrap the complaint in privacy language as if privacy-first crawl required resurrection.

What we publish and Editorial Standards already treat originality as a need test on the publisher side. That need test shrinks the public www index when honesty requires it. Privacy-first crawl does not reverse that shrink. It also does not enlarge the Search index to compensate. A company that cleaned overlapping privacy essays on www did not quietly promise that every third-party URL would appear in Search the next morning. Corrections made the publisher record longer when the company was wrong. Length of the Corrections list is not length of the Search index.

Contact’s Search-removal path assumes a result was present and asks for the exact result URL and a reason. The converse assumption — that every page you care about must already be present because Search is private — is what this essay rejects. Privacy is not a coverage coupon.

## What this essay will not claim

It will not claim Oernoe Search indexes the whole web. It will not claim privacy-first crawling is empty marketing; the guide’s data-diet sentences are real limits. It will not claim Status green enlarges crawl coverage. It will not invent crawl percentages, user counts, or fake index sizes. It will not walk Search chrome control by control. It will not claim unreleased names are live homepage products. It will not turn every empty result into a Corrections entry. It will not clone the rate-limits essay, the finite-crawl essay, or the ordinary-missing essay under a new title.

## The useful sentence

Privacy-first crawl is about what identifying data the fetch refuses to store. Infinite index is a fantasy about how large the store becomes. How Search works lists the first. Search’s tip footer denies the second. Keep them separate, and privacy language stops pretending it can conjure coverage.