What the shared tracker measures.
On a participating website, the tracker records an anonymous browser-session visit and the first observed referral source for that website and Eastern calendar day. It may also send a single boolean hint that a trusted pointer or keyboard interaction occurred on a visible, focused page. It does not send the key, typed text, pointer position or element clicked.
A browser session counts at most once per participating website per America/New_York calendar day. Different browsers, new cookies and returns on later days can count separately. These are estimates of browser activity, not identified people or verified attention.
A separate cookie on each website.
The new shared tracker uses the first-party website_analytics_session cookie on the website being visited. It contains a random value, uses the whole-site path, SameSite=Lax and Secure on HTTPS, and has a rolling twenty-four-hour lifetime renewed when the tracker loads. It is readable by that site’s JavaScript so the tracker can send its anonymous value; it is not an HttpOnly cookie.
Each website has its own cookie. The tracker does not copy that value between websites or use it to build a cross-site visitor profile. Clearing or blocking the cookie can prevent collection or cause a later visit to count as a new session.
Where requests go.
The script is served over HTTPS from didgasadventure.com. Participating pages send small HTTPS requests to the central /portfolio-analytics.php collector on that domain. Requests omit cross-site credentials; they do not send the owner’s login cookie or other websites’ account cookies to that collector. Only registered website origins are accepted.
The random site-cookie value and a temporary page nonce are processed transiently. Private storage retains keyed daily digests for counting and duplicate-request protection, rather than raw random identifiers. The stored data is separated by website.
Sources, categories and retention.
Source data keeps a normalized referral domain and, when supplied, short source, medium and campaign labels of up to forty-eight permitted characters. Complete referral URLs, page URLs, paths and arbitrary query strings are not stored by this feature. An entry is recorded only as a broad category such as home or content. Missing or blocked referrals can appear as direct or unknown.
Known-bot signals, declared automation, likely-human interactions and unverified activity have separate categories. A user-agent string may be checked transiently for a bot signal, but its raw value is not retained. These signals can be wrong or imitated; a likely-human figure is not proof of a person. Requests that never reach the collector are outside the report.
Aggregate reports retain up to ninety days. Anonymous daily deduplication and replay-protection digests retain up to thirty days. Collection starts when a site’s installation begins receiving events; earlier visits are not backfilled or newly classified. Private owner reports show the available dates.
What this feature does not collect.
The analytics feature stores no raw IP addresses, raw user-agent strings, raw session tokens, complete page or referral URLs, form values, passwords, payment details, typed messages, recordings, microphone audio or screenshots. It does not infer a visitor’s location. Recognized private account, login, checkout and editing routes and embedded frames are excluded; interaction hints ignore forms and editable controls.
This describes the analytics feature, not every function of every participating website. Account, purchase, household, media and other application features have their own policies. Hosting providers may separately maintain ordinary access logs.
Keep your own visits out of the report.
The owner’s private website dashboard offers a Visit without tracking link for each participating website. Opening a site with ?analytics_exclude=1 enables a per-site browser exclusion for this tracker. The host-only website_analytics_excluded cookie stores only the boolean exclusion preference and lasts up to one year. ?analytics_exclude=0 removes that preference. Use a separate choice for each website and browser; signing in at Didga does not automatically exclude activity on unrelated domains.
For a temporary test, ?analytics_test=1 excludes this tracker in that website’s current tab session. Recognized private routes and pages explicitly marked for exclusion are also skipped. These choices affect the new shared tracker; they do not reset existing public visitor charts or change a site’s separate account or application behavior.
Who can read the reports.
Detailed aggregate reports require the existing Didga owner login. Other participating websites do not receive copies of its password or owner-account files, and their public visitors cannot fetch the private report through the collection endpoint. Data and authentication secrets stay outside the public website directory.
Didga’s existing game analytics.
Didga’s Adventure keeps its already installed visitor and game-activity system rather than loading a second shared tracker. Its separate didga_visitor_session cookie is HttpOnly and its existing public chart retains up to 365 days of aggregate daily counts. Its private game report estimates game-start sessions and active play time. Other websites receive website-level visits, referrals and interaction estimates only when their tracker is installed and active; game-start and play-time adapters are not added to those games. The registered application integrations are installed, but collection still depends on the tracker running on an eligible public page and its requests reaching the collector. An empty report does not establish that nobody visited.
Existing public counters on other websites retain their own cookies and histories. The new private report neither replaces those charts nor combines them into a count of unique people. Read Didga’s recording and game privacy page for its local recording controls and existing game-specific analytics.
Coverage has practical limits.
Disabled JavaScript, cookie restrictions, blocked requests, missing referral information, shared browsers and automation can affect measurements. An available report is an observation of received signals, not a complete census of visitors. The participating-site registry deliberately excludes A-starting domains and unbuilt or parked sites. Sites that were unreachable during the inventory were left unchanged because their content could not be verified.