Deep dives into tools,
techniques & the web.
Practical guides covering developer tools, health metrics, SEO, converters and more — written to be actually useful.
✦ AI-powered — understands natural language
Why You Should Never Remove a 301 Redirect — Maintenance, Chain Accumulation, and Latency Impact Over Time
Why Major Organizations Accidentally Let Their Domains Expire — and the Monitoring That Prevents It
Domain expiry is predictable months in advance — yet Microsoft Hotmail UK, Foursquare, and dozens of others have accidentally let critical domains expire. Here's the four root causes of accidental expiry (expired payment method, abandoned monitoring email, administrative responsibility gap, forgotten legacy domain), the redemption timeline that provides recovery opportunity, and why renewal notifications should never go to the email domain that's expiring.
AI-Generated Content Has a Whitespace Fingerprint — Why Normalisation Matters Before Publishing or Processing
AI-generated content carries whitespace patterns — double-line-spacing from training data, inconsistent bullet markers, occasional zero-width Unicode characters — that cause rendering issues when pasted into CMSs. Here's how to normalise whitespace before LLM input (reducing tokens and improving output), why whitespace sensitivity in text diff changes what differences are shown, and the specific Word document whitespace artifacts (non-breaking hyphens, soft hyphens, figure spaces) that appear on export.
The "238 wpm" Average Reading Speed Describes No One Precisely — What Reading Time Estimates Actually Mean
The 238 wpm reading speed used in "8 minute read" estimates comes from a 2019 meta-analysis of college students reading non-fiction prose — it doesn't apply to technical documentation (75-150 wpm), fiction (300-400 wpm), or scanning (700+ wpm). Here's the speed-comprehension trade-off that makes speed reading programs ineffective, how Medium adjusts reading time for images, and why word count reading time estimates are calibrated for English and overestimate reading speed for German while underestimating it for Chinese characters.
News Sitemaps Are Not Standard XML Sitemaps — The Format, 48-Hour Rule, and Google News Inclusion Process Explained
News sitemaps use a different XML namespace with required fields (news:publication, news:publication_date, news:title) and a 48-hour recency rule — Google's News crawler processes them separately from standard Googlebot indexing. Here's the exact XML structure, why the 48-hour limit exists, how Google News inclusion differs from organic indexing (different system, different requirements), and what happened to the AMP requirement for Top Stories after June 2021.
Processing Speed Peaks at 25, Vocabulary at 65 — Why "Peak Age" Depends Entirely on What You're Measuring
Cognitive processing speed peaks at 18-25, but vocabulary peaks in the late 60s, emotional regulation improves into the 70s, and crystallized intelligence doesn't decline until the 70s-80s. Here's the multiple-peaks framework for human performance, why processing speed loss doesn't translate to overall cognitive decline (strategy, knowledge, metacognition compensate), why marathon runners peak later than sprinters, and the relative age effect that makes September-born UK children more likely to be identified as gifted.
Why the Sun and Moon Appear Identical in Size — Solid Angles, Steradians, and Angular Measurement in Astronomy and Optics
The Sun and Moon appear the same angular size because the Sun is ~400× larger but ~400× farther away — angular diameter, not physical size, determines apparent size. Here's solid angle (steradians) as the 3D extension of plane angles, why arcseconds are the fundamental astronomical unit (parallax, telescope resolution, atmospheric seeing), how Apple's Retina display threshold is defined in arcminutes, and why right ascension uses hours not degrees.
Why a 24 MP Full-Frame Camera Beats a 48 MP Smartphone — Area Measurement From Pixel Density to Body Surface Area
Pixel density, camera sensor size, semiconductor die area, and human body surface area are all area measurements at different scales — from nanometres² to m². Here's why a 24 MP full-frame camera outperforms a 48 MP smartphone camera in low light (pixel area and light collection), why Apple's M1 Ultra fits 88 million transistors per mm², why BSA in m² matters for burn assessment and chemotherapy dosing, and why a "1-inch" camera sensor isn't actually 1 inch.
A BIMI Logo Doesn't Stop Phishing — What BIMI's Security Model Actually Protects Against
BIMI's logo signals trademark ownership and DMARC enforcement — not email quality or sender reputation. It protects against spoofing the exact registered domain, not against lookalike domains (examp1e.com). Here's BIMI's four-layer authentication stack context, what VMC vs CMC actually certifies (trademark ownership, not email quality), why DMARC p=reject is the real security benefit in the BIMI requirement chain, and why major brands remain the primary BIMI adopters.
Why "9 AM London, Every Monday" Arrives at 3 AM in New York Some Weeks — DST, Offset Arithmetic, and Scheduling Global Teams
Tokyo doesn't observe DST; New York does — which means a weekly meeting at "9 AM London" shifts by an hour relative to Tokyo twice a year, while New York shifts at slightly different times than London does. Here's the full offset arithmetic for a London-New York-Tokyo meeting across the year, why "9 AM UTC" is more stable than "9 AM London" for global recurring meetings, and why calendar events stored with UTC offsets break at DST transitions while IANA timezone identifiers don't.
ISO 8601, Unix Timestamps, and the "Local Time in Database" Bug — A Practical Format Selection Guide
ISO 8601 UTC and Unix timestamps represent the same moments but encode different assumptions — and the most pervasive timestamp bug is storing DATETIME without timezone information, which breaks when servers change timezones, when DST transitions create missing hours, or when users in different timezones submit local timestamps. Here's the practical guide: when to use each format, why ISO 8601 UTC sorts lexicographically, and the date-only vs datetime semantic distinction that causes birthday-display bugs.
Why OAuth redirect_uri Errors Are Almost Always Encoding Problems — and How to Fix Them
OAuth redirect_uri errors are almost always URL encoding mismatches — the encoded URI sent in the authorization request must exactly match the registered URI, character for character. Here's the correct encoding for redirect_uri, why Base64 state parameters break CSRF checks when + decodes as space, the five different array encoding formats that different frameworks use, and the URL parser inconsistencies that enable SSRF attacks.
Why Forwarded Emails Inflate Campaign Metrics — UTM Attribution Edge Cases You Need to Know
A forwarded marketing email carries its UTM parameters intact — attributing the new recipient's session to the original campaign, silently inflating email metrics. Here's the forwarding attribution problem, how Facebook's fbclid and Google's gclid travel with shared URLs attributing shared clicks to paid ads, the cross-subdomain UTM tracking gap that loses attribution at subdomain boundaries, and why SPAs need to capture UTMs before client-side routing removes them.
