Manifest position 983 · CrUX rank bucket 1000
https://www.t-online.de
Coverage complete
Atomic audit completed across home, section, article, and search journey. Search returned a directly observed 403 failure state.
All 58 atomic check outcomes
| Principle / check | Status | Confidence | Evidence or reason |
|---|---|---|---|
respect-user-preferencesrespects-color-scheme | issues | high | Desktop screenshots under default and prefers-color-scheme: dark are byte-identical; computed html color-scheme is normal, so the preference does not change the experience. F01 medium: The site imposes one dark palette instead of following the user’s light/dark preference. |
respect-user-preferencesrespects-reduced-motion | pass | high | The reduced-motion probe matched the preference and found only three 0.01ms animations, effectively suppressing non-essential motion. |
respect-user-preferencesrespects-contrast | pass | high | The forced-colors/prefers-contrast screenshot preserves visible text, borders, controls, and link differentiation throughout the consent state. |
implement-natural-interactionsview-transitions | issues | high | The rendered probe found no view-transition CSS or view-transition meta declaration on the home template, while navigation is through full document links. F02 low: Navigation changes are abrupt despite the site being a multi-page reading experience. |
implement-natural-interactionsscroll-driven-animations | not-applicable | high | No scroll-linked decorative animation was observed on the sampled news templates, so there is no scroll-event implementation to assess. |
implement-natural-interactionsphysical-gestures | pass | high | Sampled navigation uses native links, buttons, and browser scrolling; no custom pointermove gesture layer or platform-fighting interaction was observed. |
provide-guided-navigationscroll-state-aware-chrome | issues | high | The 18,000px home document and article template show fixed navigation/breadcrumbs, but the CSS probe found no scroll timeline and the sampled first viewport exposes no progress cue. F03 low: Long pages provide no visible reading progress or scroll-responsive chrome. |
provide-guided-navigationanchored-positioning | not-applicable | high | No tooltip or edge-positioned transient overlay requiring anchored positioning appeared on the sampled primary paths. |
provide-guided-navigationdirects-attention | pass | high | Article breadcrumbs, descriptive navigation labels, skip links, and a single visible H1 give clear location and focus cues. |
maximize-content-reduce-noiseno-intrusive-interruptions | issues | high | Desktop and 360px screenshots show the consent layer occupying the entire viewport before interaction, with the underlying news content dimmed and inaccessible. F04 high: A full-viewport consent/paywall choice obscures all content on first load. |
maximize-content-reduce-noisesemantic-dismissible-primitives | issues | high | The DOM reports one role=dialog but zero <dialog>, popover, or details elements; the first viewport has no visible close or reject action. F05 medium: The blocking consent overlay is custom rather than a native dialog and has no clear light-dismiss path. |
maximize-content-reduce-noisereduced-chrome | issues | high | The initial desktop screenshot is almost entirely consent UI, while the no-JS article screenshot devotes roughly 210px of a 493px viewport to two navigation rows and breadcrumbs before content. F06 medium: Navigation and consent chrome dominate the first viewport. |
adapt-to-the-form-factorresponsive-no-horizontal-scroll | pass | high | Home, section, and article layouts all report scrollWidth=clientWidth=360 with viewport meta present and zero horizontal overflow. |
adapt-to-the-form-factorcomponent-level-responsiveness | pass | high | The same navigation and content-card templates reflow from desktop to 360px without overflow, while article and section layouts remain usable. |
adapt-to-the-form-factorinput-modality-aware | issues | high | The keyboard probe found 23 of the first 40 focusable elements under 24px in one dimension; the search input computed outline is none when focused. F07 medium: Many controls are too small and the search input lacks a visible focus indicator. |
support-core-task-successclear-purpose-and-primary-action | pass | high | The home page has descriptive metadata and “t-online - Anständig. Informiert.” H1; article title, section breadcrumbs, and content hierarchy clearly communicate purpose. |
support-core-task-successprimary-flow-completion | issues | high | Submitting the site’s documented search target with q=Bundestag loaded https://suche.t-online.de/web?sr=1&q=Bundestag and returned only “403 Forbidden”. F08 high: The primary site search journey returns a bare 403 page in the audited browser. |
support-core-task-successclear-system-state-and-recovery | issues | high | The search screenshot and DOM contain only “403 Forbidden”, with no navigation, recovery control, or retained query. F09 high: The failed search state offers no explanation, retry, or route back to content. |
be-fast-and-stablegood-core-web-vitals | issues | high | Lighthouse measured LCP 7.8s, interactive 12.9s, and performance 0.65; the independent trace measured LCP 5.42s. F10 high: Largest Contentful Paint is far outside the good range. |
be-fast-and-stablevisual-stability | pass | high | Layout probes measured CLS 0 on home/article and 0.000035 on section; Lighthouse measured CLS 0. |
be-fast-and-stableefficient-main-thread | issues | high | Lighthouse measured 280ms TBT; the trace captured two long tasks up to 88.53ms with 64.64ms trace TBT. F11 medium: Main-thread scripting adds noticeable blocking work. |
be-fast-and-stableefficient-resource-delivery | issues | high | HAR recorded 124 requests, 2.02MB transferred, 50 scripts, and parser-inserted high-priority consent resources; LCP was 5.42–7.8s. F12 high: The home load is heavy and critical delivery competes with consent and widget resources. |
be-fast-and-stabletrim-unused-and-duplicate-code | issues | high | Lighthouse estimates 314KiB unused JavaScript and 49KiB unused CSS, including consent, GTM, game-widget, and core advertising code. F13 medium: Substantial unused JavaScript and CSS ships on the home page. |
be-inclusivenames-roles-labels | issues | high | Lighthouse accessibility scored 0.87 and link-name failed for video and game links; DOM probes show search input labels are empty. F14 high: Several links have no accessible name and search fields lack associated labels. |
be-inclusivesufficient-contrast | issues | high | Lighthouse measured 3.59:1 for magenta #e20074 text on #f5e198, below the required 4.5:1 for the sampled 18px bold label. F15 medium: Breaking-news text does not meet WCAG minimum contrast. |
be-inclusivestructure-and-focus | issues | high | Lighthouse heading-order failed on an H3 teaser that skips hierarchy; the focus probe computed outline:none for the search input. F16 medium: Heading order is invalid and one key input has no visible focus ring. |
be-inclusivelegible-text | pass | high | Desktop, mobile, forced-color, and no-JS screenshots show readable type sizes, line wrapping, and unclipped editorial headings. |
be-inclusivezoom-reflow-targets-and-media | issues | high | All three 360px layout probes had zero overflow, but the keyboard/geometry probe counted 23 of the first 40 focusable elements under 24px in one dimension. F17 medium: Reflow works, but many targets remain undersized for touch and zoomed use. |
follow-best-practicesno-console-errors | pass | high | Lighthouse errors-in-console passed with no logged items. |
follow-best-practicessound-document-and-assets | issues | high | The image primitive found 48/100 sampled images without width/height, 17 without srcset, and 52 legacy-format assets; Lighthouse also failed unsized-images. F18 medium: Images are not consistently dimensioned or efficiently described. |
follow-best-practicesbrowser-platform-hygiene | pass | high | Lighthouse best-practices scored 1.0; notification permission remained default and no prompt appeared on load. |
be-discoverabletitle-and-description | pass | high | Home and article both expose unique descriptive titles and meta descriptions in raw and rendered HTML. |
be-discoverablecrawlable-and-mobile-friendly | pass | high | Lighthouse SEO scored 1.0, crawlable anchors passed, viewport meta is present, and robots allows normal public crawling. |
be-discoverablecanonical-and-indexing-signals | issues | high | GET /sitemap.xml returned a normal HTML page rather than XML sitemap data, while robots.txt contains no Sitemap directive; canonical URLs and HTTP 200 otherwise pass. F19 low: The conventional sitemap endpoint does not return a sitemap and robots.txt declares none. |
be-discoverablestructured-and-shareable-metadata | pass | high | Home exposes WebPage/SearchAction JSON-LD and Open Graph; the article exposes article JSON-LD aligned with its visible H1 and canonical URL. |
be-private-and-securesecure-transport-and-headers | issues | high | HTTPS, HSTS, nosniff, and CSP are present, but CSP is only “frame-ancestors self https://*.t-online.de” and Permissions-Policy is absent. F20 medium: Transport is strong, but the CSP only controls framing and does not constrain scripts or other content. |
be-private-and-securedata-minimisation-and-third-parties | issues | high | HAR recorded 62 third-party requests and 1.12MB third-party transfer (56% of bytes); tracker evidence found 10 third-party origins and Google Tag Manager. F21 high: The page loads a large third-party and tracking footprint. |
be-private-and-securein-context-permissions-and-modern-auth | pass | high | No notification prompt was fired on load and notification permission remained default; the sampled public reading paths do not require authentication. |
be-private-and-securedefensive-browser-policies | issues | high | HSTS and CSP frame-ancestors protect transport/clickjacking, but Permissions-Policy is absent and CSP provides no script-src/Trusted Types defense. F22 medium: Browser-enforced policy coverage is incomplete. |
be-resilientprogressive-enhancement | pass | high | Discoverability measured 99% raw-HTML content coverage on home and 100% on article, with title, H1, and descriptions available without JavaScript. |
be-resilientresilient-runtime-behaviour | pass | high | The consent interface fits at 360px and forced colors, native navigation remains available, and sampled layouts have no clipping or horizontal overflow. |
be-resilientoffline-and-installable | issues | high | A manifest link exists, but the fetched manifest parsed as an empty object; the browser found zero service-worker registrations and zero caches. F23 medium: The public news experience is neither installable nor prepared for offline use. |
be-resilientnetwork-and-http-failure-states | issues | high | The representative search request returned only a bare 403 page, with no retry, navigation, or explanation. F24 high: HTTP failure handling collapses to an unbranded blank error. |
be-internationalisedlang-dir-and-logical-properties | pass | high | All sampled public content declares html lang=de; German reading order, umlauts, and layout render correctly at desktop and narrow widths. |
be-internationalisedlocale-aware-data | pass | high | Rendered German financial data uses locale-appropriate decimal commas (for example DAX +0,61%) and German text/date conventions across templates. |
be-internationalisedtime-zone-correctness | not-applicable | high | The sampled news, section, and search surfaces exposed no scheduling or cross-time-zone workflow to test. |
be-trustworthyno-dark-patterns | issues | high | Both desktop and mobile first-load screenshots show “ZUSTIMMEN” and “JETZT SICHERN” as the only primary actions; rejection requires entering the privacy manager. F25 high: The consent wall presents acceptance and a paid subscription as the prominent choices, without an equal reject option. |
be-trustworthyhumane-error-handling | not-applicable | high | The sampled forms have no required-field validation flow; search submission navigates directly and the resulting HTTP failure is judged under task success and resilience. |
be-trustworthytrustworthy-input-assistance | issues | high | The home form probe reports name=email with autocomplete empty; search explicitly uses autocomplete=off. F26 low: The email sign-in field omits autocomplete metadata. |
be-trustworthysafe-commercial-and-account-flows | issues | high | The blocking dialog places “Kostenfrei mit Werbung” beside a €2.99/month PUR subscription while omitting an equally prominent reject path. F27 medium: The first-load commercial/consent choice conflates privacy consent with a PUR subscription offer. |
be-sustainableoptimised-assets | issues | high | The image audit found 52/100 legacy-format images, 17 missing responsive srcset, 12 below-fold images not lazy-loaded, and 48 missing dimensions. F28 medium: Asset delivery leaves avoidable image waste. |
be-sustainableno-wasteful-work | issues | high | Lighthouse estimates 314KiB unused JavaScript and HAR shows 50 script requests; 62 third-party requests account for 1.12MB. F29 medium: Unused and third-party work consumes substantial transfer and execution. |
be-sustainablethird-party-and-media-budget | issues | high | HAR attributes 1,123,471 of 2,017,478 transferred bytes to third parties, led by consent and engagement widgets. F30 high: Third parties consume more than half of the initial transfer budget. |
be-agent-readystructured-agent-capabilities | not-applicable | high | This public editorial site exposes content rather than an agent-invocable transaction surface; emerging WebMCP capability is not required for the sampled scope. |
be-agent-readyon-device-inference | not-applicable | high | No user task in the sampled editorial and search surfaces requires on-device inference; absence is not a quality failure. |
be-memory-efficientno-leak-under-repeated-interaction | pass | high | After ten full-page scroll cycles, heap snapshot self-size rose only 86,323 bytes (0.22%); DOM nodes and animation counts were unchanged. |
be-memory-efficientbounded-footprint | pass | high | The hydrated page used about 28–29MB JS heap with 5,592 DOM elements; the 39.4MB snapshot footprint is proportionate to a large news index. |
be-memory-efficientno-detached-dom-or-unbounded-listeners | pass | high | Repeated scrolling left DOM nodes fixed at 5,592 and animations fixed at 11; constructor populations and closures were effectively stable between summaries. |
Provenance
Canonical report: results/atomic/reports/0983-www_t-online_de.json
Report SHA-256: ad25e929999b1a26f87bd2927faee6dc95d25b9ad9e1f6d9943e62c00e927d04
Local retained report: runs/2026-07-17T17-27-24-856Z/atomic-reports/www_t-online_de/2026-07-24T07-44-44-407Z/report.json
Local evidence root: runs/2026-07-17T17-27-24-856Z/atomic-reports/www_t-online_de/2026-07-24T07-44-44-407Z
Catalog SHA-256: sha256:78ccfdb2d483f4c57d9dafed80fd86c6265585a56457c8dcfddc254b80fb44d7
Raw screenshots, HARs, traces, heaps, and other browser artifacts are retained at the local evidence root and intentionally are not committed. Artifact paths in the canonical report are relative to that root.