Search the archive

Full-text search across article titles and summaries. Refine with sector, year, content type, and reading-time filters.

Results

29,182 matches

POSTGRESQL BLOAT IS A FEATURE, NOT A BUG

PostgreSQL bloat is a fundamental design feature where databases grow and queries slow down due to the accumulation of old data versions. This occurs because the database's Multi-Version Concurrency Control (MVCC) creates new physical tuples for every update and only marks old ones as obsolete upon deletion, rather than overwriting or erasing them. These "dead tuples" consume disk space, force the system to perform unnecessary I/O by scanning full pages, and contribute to index bloat.

2026.02.19
rogerwelin.github.io14 min readTLDR Dev

VIRTUAL SCROLLING FOR BILLIONS OF ROWS — TECHNIQUES FROM HIGHTABLE

HighTable is a React component that lets you scroll through billions of rows in a browser table smoothly. It only loads and renders the ~30 rows you can actually see at any given time. The tricky part is that browsers cap element heights at ~17M pixels, so the creative solution is to use a downscaled scrollbar plus a dual scrolling mode (small gestures scroll locally with pixel precision, big ones jump globally).

2026.02.19
rednegra.net30 min readTLDR Dev

WE RALPH WIGGUMED WEBSTREAMS TO MAKE THEM 10X FASTER

Vercel found that WebStreams were a huge bottleneck in Next.js server rendering. The streams were burning CPU on Promise allocations for every chunk. Vercel built fast-webstreams, a drop-in replacement that collapses piping chains into a single Node.js pipeline() call with zero Promises per chunk, with the end goal of merging this package into Node.js itself soon.

2026.02.19
vercel.com9 min readTLDR Dev

WHEN INTERFACES BECOME DISPOSABLE

A new parent used AI coding to create a custom interface for his Fitbit data, addressing his unique sleep tracking needs that the standard app couldn't fulfill. This bespoke tool is an example of disposable software. AI agents are increasingly enabling users to bypass traditional interfaces and directly access product capabilities via APIs, creating a fragmented landscape where custom solutions are easily generated.

2026.02.19
chrisloy.dev9 min readTLDR Dev

WHAT IS HAPPENING TO WRITING?

There's widespread acceptance of AI-generated slop in writing. Despite the allure of AI's efficiency, genuine writing is an irreplaceable process of solitary thought, labor, and public engagement. It's good to be aware of cognitive debt from over-reliance on AI.

2026.02.19
resobscura.substack.com19 min readTLDR Dev

REACT DOCTOR

React Doctor is an open-source tool designed to help coding agents diagnose and fix issues in React codebases. It scans projects for security, performance, correctness, and architecture problems, outputting a 0-100 score with actionable diagnostics.

2026.02.19
github.comGitHub repositoryTLDR Dev

ELECTROBUN V1

Electrobun v1 is a new desktop application framework for building ultra-fast, tiny, and cross-platform apps with TypeScript. It has first-class support for macOS, Windows, and Ubuntu, providing automatic installers, robust auto-updates, and advanced features like true process-isolated webviews (OOPIF).

2026.02.19
blackboard.sh3 min readTLDR Dev

SLOWMO

Slowmo is a tool designed to universally control the time of web content, allowing users to slow down, pause, or speed up animations and other time-based elements. It has purposes such as debugging animations, studying interactive web demos, and adjusting the difficulty of web games.

2026.02.19
github.comGitHub repositoryTLDR Dev

GRANDPA LISSAJOUS – A 13-AGENT AI ORCHESTRATION LOOP

The Grandpa Loop uses 13 Simpsons-themed AI agents to automate software development through non-linear feedback loops. It has adversarial spec reviews, automated builds, and manual-style UX testing. A central observer agent monitors performance data to tune the pipeline, making sure the system self-corrects and improves cycle times.

2026.02.19
blog.sauhsoj.wtf6 min readTLDR Dev

YOU ONLY THINK THEY WORK FOR YOU

External vendors prioritize industry relationships over individual clients. To maximize value, leaders should treat consultants as educators to gain strategic knowledge. This shift is especially important as AI agents begin automating PR and distribution, potentially creating "black box" processes that muddy operational insights.

2026.02.19
steveblank.substack.com5 min readTLDR Dev

SPELL CHECKING A YEAR'S WORTH OF HACKER NEWS

Using Claude Haiku 4.5 and Gemini 3 Flash, the author identified 1,430 errors across 7,000 Hacker News blogs. After manual validation, 404 emails were sent, with a 50% response rate. This is an example of AI's potential for automated kindness and the need for human-in-the-loop verification.

2026.02.19
fi-le.net12 min readTLDR Dev

COSMOLOGICALLY UNIQUE IDS

Probabilistic random IDs offer a practical and efficient solution with negligible collision risk, while deterministic hierarchical schemes, despite guaranteeing absolute uniqueness, become impractically large for a universe-spanning civilization.

2026.02.19
jasonfantl.com25 min readTLDR Dev

HOW AI IS AFFECTING PRODUCTIVITY AND JOBS IN EUROPE

AI adoption increases labor productivity by 4% without short-term job losses, but its benefits are unevenly distributed, favoring larger firms and those investing in complementary assets like human capital.

2026.02.19
cepr.org13 min readTLDR Dev

Sponsored

OPERATIONALIZED

Attackers don't go through your tools — they go around them, chaining identity relationships to reach critical assets.

2026.02.19
specterops.ioSponsorTLDR InfoSec

SUPPLY CHAIN ATTACK EMBEDS MALWARE IN ANDROID DEVICES

Keenadu is a firmware-level Android malware embedded via supply chain compromise that hijacks Android's Zygote process to inject itself into every app on infected devices. It affects 13,000 devices across Russia, Japan, Germany, Brazil, and the Netherlands. Operating as a multistage loader, Keenadu deploys modules targeting shopping platforms such as Amazon and Temu, monitors Chrome queries, and commits ad fraud. Organizations should cross-reference Kaspersky's published IoCs against managed Android fleets. Firmware-level infections require full firmware replacement, and system app infections are mitigated by disabling or replacing the compromised app.

2026.02.19
links.tldrnewsletter.com3 min readTLDR InfoSec

SHINYHUNTERS ALLEGEDLY DROVE OFF WITH 1.7M CARGURUS RECORDS

Cybercrime gang ShinyHunters has claimed the theft of 1.7 million CarGurus corporate records, including personal and internal data, and is extorting the company with a February 20 deadline. The crew has chalked up 15 alleged breaches in 2026, hitting financial firms, retail brands, car marketplaces, and dating platforms, often via social engineering and SSO abuse.

2026.02.19
theregister.com2 min readTLDR InfoSec

STOP THE CAP: MAKING ENTRA ID CONDITIONAL ACCESS MAKE SENSE OFFLINE

CAPSlock is an offline Conditional Access Policy (CAP) analysis engine built on ROADrecon that simulates sign-in scenarios without generating tenant artifacts and distinguishes between definitively applied policies and signal-dependent ones. The tool addresses a critical gap in Entra ID assessments, overlapping CAPs create hard-to-reason enforcement behavior, particularly around MFA bypasses and platform-specific policy gaps, by modeling normalization, exclusion logic, and cumulative policy enforcement offline. Both red teamers and defenders should evaluate CAPSlock for CAP auditing workflows, with the caveat that pending AAD Graph deprecation will require elevated Microsoft Graph permissions for future policy collection.

2026.02.19
specterops.io18 min readTLDR InfoSec

CLICKFIX: STOPPED AT ⌘+V

This post details a lightweight macOS defense against ClickFix attacks — now integrated into BlockBlock v2.3.0 — that intercepts ⌘+V keystrokes via NSEvent global monitoring, pauses the terminal process with SIGSTOP, and prompts the user to review clipboard contents before execution. The approach works across major terminal emulators and is particularly effective given that most documented ClickFix campaigns — including recent campaigns by North Korean APT UNC1069 and LLM-propagated attacks — explicitly instruct victims to use the keyboard shortcut. Notable limitations include no coverage for right-click paste operations, a dependency on Accessibility permissions, and Apple's Endpoint Security framework lacking an equivalent ES_EVENT_TYPE_AUTH_PASTE event, making user-space interception the only practical approach.

2026.02.19
objective-see.org7 min readTLDR InfoSec

THE CURIOUS CASE OF THE TRITON MALWARE FORK

A suspicious fork of the Triton macOS omg.lol client turns out not to be simple plagiarism, but a clumsy malware delivery attempt hosted on GitHub. The attacker floods the README with links to a Windows-only malware ZIP hidden in an Xcode asset path, then pads the GitHub contribution graph with fake backdated commits to appear legitimate. Sandbox analysis shows a multistage Windows infection chain using 7zip, LuaJIT, anti-analysis tricks, and C2 traffic masquerading as Microsoft and blockchain traffic.

2026.02.19
brennan.day13 min readTLDR InfoSec

ADWSDOMAINDUMP

ADWSDomainDump is an Active Directory domain dumper that uses ADWS (port 9389) instead of LDAP to enumerate AD objects, bypassing both Microsoft Defender for Endpoint and CrowdStrike Falcon detection.

2026.02.19
github.comGitHub repositoryTLDR InfoSec

NONO

Nono is a secure, kernel-enforced sandbox for AI agents, MCP, and LLM workloads. It features capability-based isolation with secure key management and blocking of destructive actions in a zero-trust environment.

2026.02.19
github.comGitHub repositoryTLDR InfoSec

COGENT SECURITY (PRODUCT LAUNCH)

Cogent Security is an agentic AI platform for vulnerability management that autonomously investigates, prioritizes, and orchestrates remediation of security issues.

2026.02.19
cogent.comTLDR InfoSec

PASSWORD MANAGERS LESS SECURE THAN PROMISED

ETH Zurich's Applied Cryptography Group demonstrated 25 total attacks across Bitwarden (12), LastPass (7), and Dashlane (6), collectively serving ~60 million users breaking their "zero-knowledge encryption" guarantees by operating a malicious server that deviates from expected client behavior during routine actions such as login, vault access, and sync. Attacks ranged from targeted vault integrity violations to full organizational vault compromise, requiring only lightweight server-impersonation tooling rather than significant computational resources, with the root cause attributed to feature-bloat complexity and reliance on obsolete 1990s-era cryptographic primitives. Security teams should evaluate password manager vendors against criteria of transparent vulnerability disclosure, third-party audits, and modern end-to-end encryption enabled by default, pending publication of full findings at USENIX Security 2026.

2026.02.19
ethz.ch3 min readTLDR InfoSec

DUTCH SECRETARY OF DEFENSE THREATENS TO 'JAILBREAK' NATION'S F-35 JET FIGHTERS

Dutch defense chief Gijs Tuinman suggests F-35 fighter software could be “jailbroken,” raising sharp questions about European dependence on US control systems and cloud services. The question is whether there are hidden kill switches or how Lockheed's update and logistics infrastructure can effectively ground fleets.

2026.02.19
tomshardware.com2 min readTLDR InfoSec

FROM BRICKSTORM TO GRIMBOLT: UNC6201 EXPLOITING A DELL RECOVERPOINT FOR VIRTUAL MACHINES ZERO-DAY

Mandiant and GTIG described how UNC6201 exploited CVE-2026-22769 (CVSS 10.0), a hardcoded admin credential in Dell RecoverPoint's Apache Tomcat configuration located at /home/kos/tomcat9/tomcat-users.xml. This enabled unauthenticated deployment of WAR files via/manager/text/deploy and allowed root-level code execution. Persistence was maintained by hijacking convert_hosts.sh, which is run at boot via rc.local. GRIMBOLT, a C# backdoor replacing BRICKSTORM in September 2025, is compiled with native AOT to remove CIL metadata, making static analysis more difficult. It shares C2 infrastructure with its predecessor and introduces new lateral movement techniques in VMware, such as using "Ghost NICs" and iptables-based Single Packet Authorization on compromised vCenter appliances. Defenders should review /home/kos/auditlog/fapi_cl_audit_log.log for requests to /manager, search for GRIMBOLT using published YARA rules and IOCs, immediately apply Dell's patch, and monitor for changes to convert_hosts.sh.

2026.02.19
cloud.google.com11 min readTLDR InfoSec

Sponsored

DECISIONS

By the time insights reach execution, market windows close. A.Team builds intelligence systems that COMPRESS INSIGHT-TO-ACTION FROM WEEKS TO HOURS using the data you already have, embedded directly in your workflows.

2026.02.19
a.teamSponsorTLDR Marketing

CREATIVE TRENDS FOR REDDIT-FIRST CAMPAIGNS IN 2026

Reddit campaigns are focusing on 4 trends. Nostalgia-driven storytelling connects through shared memories and early internet aesthetics using imperfect visuals. User-generated social proof highlights real community posts to build trust. Niche-inspired campaigns speak directly to specific passions and subcultures using their language and style. Campaigns are designed to unfold publicly, letting the community shape the story and evolve the narrative over time.

2026.02.19
cdn.prod.website-files.com7 min readTLDR Marketing

WHY 2026 MUST BE THE YEAR BRANDS BET ON THEIR OWN TEAMS

More brands are building in-house agencies as strategic assets rather than cost-saving experiments. Nearly 9 in 10 now operate internal creative teams. However, many remain underpowered because they are treated as executional support rather than full partners. Without clear operating models, proximity to the brand goes to waste. Investing in culture and empowerment allows these teams to perform at their best.

2026.02.19
links.tldrnewsletter.com4 min readTLDR Marketing

FORMULA FOR CREATING 10 B2B CAROUSELS/WEEK

Carousels continue to perform well on LinkedIn, with the front cover doing 80% of the work. Use a title that is clear, specific, and easy to skim. Combine words with images, screenshots, or emojis to signal the content and give viewers a reason to stop scrolling. Maintain a consistent structure with predictable headings and clear hierarchy. End with a question to encourage comments and interaction.

2026.02.19
linkedin.com2 min readTLDR Marketing

5 WEBINAR TIPS TO INCREASE REGISTRATIONS AND ATTENDANCE RATES

To increase show-up rates, send an email within 4 hours of the event. Test whether using “tomorrow” or the actual day in the subject line performs better. Include a bracket at the start of the subject, such as (Today!) or (Going live in 1 hour), to signal urgency. Replace the word “webinar” with alternatives like live event, digital workshop, insider session, or live panel to make the invitation feel more engaging.

2026.02.19
linkedin.com1 min readTLDR Marketing

Sponsored

WHO IS YOUR WEBSITE REALLY FOR?

If AI handles discovery and research for buyers, your site needs to work for retrieval bots, not just humans. Scrunch reveals how AI agents experience your site—what gets surfaced, ignored, and why. Get a FREE AI SITE AUDIT [10] to find content gaps and technical blockers.

2026.02.19
scrunch.comSponsorTLDR Marketing

HOW TO MINE GOOGLE SEARCH CONSOLE FOR CONVERSATION DATA

Google Search Console (GSC) can surface long conversational queries that look like AI prompts, which gives marketers a practical data source for prompt tracking even though LLM platforms like OpenAI and Google do not fully share prompt data. Some AI-related queries and AI Mode data appear in GSC, and impressions have risen over the last three months during AI feature rollouts. Use a regex filter for queries that contain 10 or more words to identify prompt-style searches that often read like detailed questions or tasks. Export these queries and analyze them in tools like Claude to uncover customer concerns, brand comparisons, price sensitivity, and recurring reputation topics.

2026.02.19
nectivdigital.com5 min readTLDR Marketing

PERPLEXITY PULLING SPONSORED ANSWERS FROM AI PLATFORM

Perplexity is phasing out sponsored answers, stepping back from one of the earliest experiments in chatbot advertising. Executives said even clearly labeled ads risk undermining user trust, which is critical to driving subscriptions and repeat usage. The move contrasts with rivals like OpenAI and Google, which are testing ads within AI responses to offset rising infrastructure costs. As AI adoption surpasses 60% among US adults, platforms are diverging on whether long-term revenue should come from ads, enterprise contracts, or paid tiers.

2026.02.19
pymnts.com2 min readTLDR Marketing