Omarchy expands AI-credit patronage with Grok tokens and Greptile PR reviews
Omacom expands in-kind AI and compute patronage and reports nearly 1,500 Greptile-reviewed PRs. Pledges are not audited cash revenue.
Search
Search results for “Cloudflare One Unified Routing”
Find published research by company, product, platform or technology.
Showing 301–320 of 394 dossiers
Google and Bing back shared AI-training opt-out — Bing adds separate grounding rules with a 2027 deadline
Publishers can opt out of AI model training without necessarily leaving traditional search or AI citations. Microsoft's new page-level grounding controls introduce a migration trap for sites relying on legacy NoArchive/NoCache settings.
Firefox plans to stagger security fixes so AI attackers cannot combine unpatched browser bugs
Mozilla reports more than seven times as many Firefox security fixes as a year earlier, while adopting weekly releases, confidential regression tests and a proposed strategy to avoid overlapping exploit-chain patch gaps.
Go servers inherit Linux TCP kernel risk through default MPTCP listeners — patch the kernel, not just Go
A Linux IPv4 TCP use-after-free can be exploitable through network-facing Go services even when their own dependencies are current. Check the running kernel and vendor patch status; a GODEBUG setting does not remove the kernel flaw.
Google Merchant API exposes AI Mode and Gemini checkout funnels — with an alpha reporting trap
Retailers and commerce-tool developers can now query first-party UCP shopping funnels rather than infer agentic checkout performance from traffic alone, but the alpha API needs careful query design.
Reddit splits its home feed into For You and Following — community discovery gets a separate front door
Reddit's two-feed home screen separates recommendation-driven discovery from subscribed communities. The change matters to publishers and moderators, but any reach uplift is still unproven.
Mistral Large 4 weights are downloadable — but hosted model services need a separate licence
The trillion-parameter Mistral Large 4 is no longer API-only; builders can self-host, but must assess licence and hardware limits.
PostgreSQL Anonymizer 3.3 patches 12 masking bypasses — upgrades require more than a version bump
PostgreSQL Anonymizer 3.3 fixes a dozen privacy and privilege flaws. Its migration is unusually sensitive: no ALTER EXTENSION UPDATE, possible salt exposure, and new restrictions on masking functions.
RevenueCat's 4,964-app study finds Black Friday offers lift sign-ups — not necessarily customer value
Holiday subscription offers correlate with more payers, but lower revenue per buyer and substantial selection bias complicate the economics. Christmas-only offers show no clear advantage.
Artificial IntelligenceWeb DevelopmentJavaScriptAI CodingBrowsers & Web StandardsDeveloper ToolsActive dossier
Google's AI bug hunter finds different V8 flaws from fuzzing — Project Zero says keep both
A six-month real-world comparison shows AI code-review agents and specialised fuzzers find substantially different V8 vulnerabilities. Replacing fuzzing with agents would leave gaps, and cost per bug remains unknown.
Tailscale 1.104.4 fixes Tailworm SSH root-escalation path — headscale needs separate review
Tailscale 1.104.4 is now tagged as a stable release after Tailworm's SSH and Serve findings. Audit SSH forwarding, update clients and assess headscale separately.
Google tells platforms to keep link-only generated pages out of Search by default
Google now advises noindex or crawl blocking for user-created pages meant for direct sharing, plus Search Console and platform-level spam monitoring.
Google is writing positive-only review summaries inside Search ads — and advertisers cannot edit them
Google Ads now documents positive-only AI review summaries inside Search ads in the US, UK, Australia and India. Advertisers cannot customise the text; source, opt-out and reporting controls remain unclear.
GitHub Copilot adds model-judged permissions to its GA agent sandbox — with unattended-run traps
Copilot now offers model-reviewed action approvals between manual and Allow All. Unlike OS sandboxing, this probabilistic gate can approve unsafe operations; unattended CLI flags may leave unexpectedly broad permissions.
Safari 27 blocks email-based identity resolvers at the network layer — first-party personalisation remains
Safari 27 blocks third-party email-to-ID resolvers by domain. First-party personalisation remains allowed; the affected vendor list and measured advertising impact remain unverified.
Artificial IntelligenceCloud & InfrastructureCompute & AI InfrastructureOpen ModelsInference & APIsActive dossier
TorchNeuron makes AWS Trainium a native PyTorch device — with substantial beta operator gaps
TorchNeuron changes the software portability equation for Trainium, making it easier to test non-NVIDIA AI accelerators without rewriting the model stack. Beta gaps still demand model-by-model verification.
NativePHP Mobile patches seven security flaws — rebuilding and app-store releases are required
NativePHP mobile 4.6.1 and 3.3.11 fix seven flaws; developers must regenerate, rebuild and distribute app updates, while older 1.x/2.x lines are unsupported.
Google will extract landing-page images for new Performance Max campaigns by default — even when accounts opt out
Google is changing the default for Performance Max image extraction on December 4, 2026. Existing campaigns keep their settings, but new campaigns will opt in regardless of an account-level opt-out; agencies and SaaS ad tools need a campaign-level override and careful handling of repeated asset settings.
Google loosens AdSense video placement rules for floating and sticky players
AdSense's October 22 policy revision changes where publishers can put accompanying video and how sticky players may start, without removing other video-ad safeguards or guaranteeing higher revenue.
Muse Wonder's API prices and 1M context are public, while Wonder Full stays inside Muse Code
The concrete API price and access limits close a key gap in the October launch: developers can cost Wonder through Model API but must not confuse maximum reasoning effort with Wonder Full.