Hacker News, Distilled

Founder Liquidity: What It Is and Who Gets It

• Many comments note that in Series A/B rounds, founders often sell a small portion of their common stock (“secondaries”) alongside new primary investment, sometimes into the low- to mid–7 figures.
• Several argue this de-risks founders’ personal lives (paying off debt, securing housing) and better aligns risk appetites with VCs, who prefer “swing for the fences” behavior.
• Critics say this is often hidden from employees, undermining the “all‑in founder” myth and creating a perception gap about who is actually still risking what.

Employees and Access to Liquidity

• It’s rare for non-founder employees to be included in early tender offers; when they are, it’s usually limited (e.g., 5–20% of vested equity, sometimes tenure-gated).
• Some think founders should only take liquidity if all employees can participate pro rata; others say founders’ unique risk and replaceability justify asymmetric treatment.
• Regulatory/tender-offer rules (e.g., limits around number of sellers) are mentioned as a practical barrier to broad employee participation.

Equity Structure: Options, Early Exercise, and 90-Day Windows

• Strong criticism of the standard 90-day post-termination exercise window; several suggest 5–10 year windows and note a small but growing list of startups doing this.
• Many advocate early exercising options and filing 83(b) elections when cheap, to avoid later AMT hits and losing equity on departure; others counter that this is risky if exercise costs or tax bills are large and liquidity is uncertain.
• Confusion is common around what 83(b) covers (unvested stock vs options) and what companies can or can’t “restrict” (early exercise vs filing itself).

Founders vs Early Employees: Risk, Reward, and Morale

• Founders emphasize years of low/no salary, personal debt, reputational risk, and inability to simply “quit,” arguing this justifies 20–50x higher ownership.
• Early employees push back that they often work similar hours, take below-market pay, and can still end up with trivial or zero outcomes—even on sizable exits—due to dilution, preferences, and opaque cap tables.
• Widespread view: being an early employee is usually a bad financial bet versus FAANG or later-stage startups; worthwhile mainly for experience, autonomy, or enjoyment, not EV.

Transparency, Ethics, and Possible Reforms

• Many see secrecy around founder liquidity and complex cap tables as exploitative; some call for more transparency tools, standardized employee-friendly terms, or even regulation.
• Others frame it as straightforward market dynamics: founders and investors will offer the minimum terms needed to hire; it’s on employees to understand and negotiate or walk away.

Overall Reaction to T‑Mobile “Lifetime” Price Lock

• Many see T‑Mobile’s “lifetime” or “no price hike” marketing as misleading once fine print is revealed (only last month’s bill covered if prices rise).
• Commenters argue this creates almost no real downside for T‑Mobile: if customers stay, they pay more; if they leave, T‑Mobile loses what they would have lost anyway.
• Several expect eventual class actions but are cynical about outcomes (small payouts, gift cards, short expirations).

Contract Fine Print, Enforcement, and Courts

• Widespread view that verbal promises from telecom sales reps are worthless without written proof; others note even written proof can be stonewalled.
• Fine print and “only certain officers can bind us” clauses are seen as tools to nullify plain-language promises.
• Some blame courts and a business‑friendly legal environment for allowing ads that conflict with buried terms.
• Others suggest this is primarily a false‑advertising issue, not just contract law.

Customer Experiences Across Carriers

• Multiple stories of price quotes from major US ISPs/mobile carriers (Verizon, T‑Mobile, AT&T) being quickly undermined by later hikes or hidden fees.
• Some users with old grandfathered plans report either:
  • Being quietly upgraded while keeping price, or
  • Eventually being forced off via network shutdowns or broken functionality.
• T‑Mobile is described by some as rigid and “robotic” in customer service; AT&T and Verizon are also criticized but occasionally preferred as the “least bad.”

Alternatives: MVNOs and Prepaid

• Many recommend MVNOs (e.g., Helium, Mint, AT&T prepaid, T‑Mobile Connect) as far cheaper for light/moderate users.
• Trade‑offs noted:
  • Network deprioritization during congestion.
  • Barebones or inconsistent customer support.
  • Limited high‑tethering or heavy‑use options.
• Some say postpaid family plans and device financing remain competitive for heavy users or multi‑line households.

Regulation, Competition, and the Sprint Merger

• Strong criticism of the T‑Mobile–Sprint merger; posters say prices rose despite promises they would fall.
• Debate over whether blocking the merger would have helped, with some arguing Sprint would have gone bankrupt anyway.
• Calls for tougher antitrust enforcement and skepticism toward merger promises about prices.

International and Policy Comparisons

• Non‑US commenters highlight much cheaper mobile data abroad, especially in Europe.
• Disagreement over whether higher US prices are justified by geography, density, and labor costs.
• Brief tangents compare telecom “grandfathering” to property‑tax caps and housing fairness.

BSP in Doom: Genius vs. Solid Application

• Many commenters note BSP trees were already standard in computer graphics and SIGGRAPH literature.
• The “genius” is framed less as invention and more as choosing the right known technique and making it work on very weak early‑90s PCs.
• Some recall BSP and related structures (octrees, sector/portal systems) being common knowledge among graphics programmers, but acknowledge implementing BSP robustly was non‑trivial and relatively rare in games.

Alternative Visibility / Partitioning Techniques

• Other engines of the era (e.g., Jedi, Build, Unreal) used sectors and portals rather than BSP, trading preprocessing cost for runtime flexibility.
• Discussion covers tradeoffs:
  • BSP: elegant, good visibility culling, but expensive preprocessing, numeric fragility, tree-walks with poor cache/branch behavior on modern CPUs/GPUs, and modding pain.
  • Portals/sectors: simpler, map‑structure‑aligned, batch‑friendly, but cause sudden visibility step-changes and can blow triangle budgets.
  • Grids/voxels and quad/octrees are mentioned as alternatives; memory/perf tradeoffs on 386‑class hardware are described as unclear.

Hardware Constraints and PC Gaming History

• Debate over whether early PCs were “meant” for games or primarily for business:
  • One side says games were common and drove home PC purchases by the early 90s.
  • The other emphasizes that PCs lacked dedicated gaming hardware (sprites, scrolling, audio) that contemporaneous home computers and consoles had.
• Doom is portrayed as a software‑driven workaround for missing graphics hardware: VGA framebuffers vs. sprite/scrolling chips.
• Discussion traces a rough transition: early 90s “multimedia” era (sound cards, SVGA, CD‑ROM), then mid‑late 90s 3D accelerators as the real turning point.

Research, Old Papers, and Not‑Invented‑Here

• Strong theme: reading older literature is a “superpower”; many breakthroughs in practice come from rediscovering 1960s–1980s work made practical by modern RAM/cache/CPUs.
• Examples include old “impossible” algorithms becoming trivial on contemporary hardware and classic routing/vision/ML methods outperforming newer ideas.
• Several complain about:
  • Paywalled ACM/IEEE papers.
  • “Write‑once‑read‑never” preprint flood, especially in ML.
  • Industry’s ahistorical tendency and NIH syndrome leading to repeated reinvention.
• Others point out that in some domains (graphics, cryptography) deep literature use is normal and modern papers add necessary formalism.

Paper Readability and Jargon

• Mixed views on whether research should use “plain English” vs. dense formalism and domain jargon.
• Older CS papers are remembered as shorter and more readable but often underspecified; newer ones are longer, more precise, and proof‑heavy but harder for non‑experts.

Tools, Books, and Learning Culture

• Commenters reminisce about foundational graphics texts (e.g., major graphics textbooks, gem collections), early compilers, and learning from books and magazines before ubiquitous open source and Stack Overflow.
• One thread laments that modern agile/SaaS environments and dopamine‑driven open source make it harder to justify days spent just reading.

Other Game‑Engine Anecdotes

• Crash Bandicoot is highlighted as an example of “inelegant but better” engineering:
  • Precomputed per‑position visibility instead of runtime VSD.
  • Per‑vertex animation with custom compression.
  • Heavy CD streaming pushing hardware beyond its nominal seek lifetime.
• There are notes on vertex animation and streaming in other 90s engines, and on early VR projects also using BSP trees.

Context and Motives for Dropping the Suit

• Many see the lawsuit as “press release lawyering” that was unlikely to succeed legally.
• Several comments suggest Musk withdrew to avoid damaging legal discovery and depositions, especially after unfavorable emails were released.
• Others think he mainly wanted to cast public doubt on OpenAI’s for‑profit shift and then moved on once that PR goal was achieved.

OpenAI’s Nonprofit-to-For-profit Shift

• Strong criticism that OpenAI abandoned its original open, nonprofit mission and is now a typical profit‑driven big‑tech entity.
• Counterpoint: structural reasons for spinning out a for‑profit arm are noted (tax rules, commercial activity), with comparisons to Mozilla/Wikimedia.
• Dispute over causality:
  • One side argues Musk pulled a large funding pledge when he was denied control, forcing OpenAI toward Microsoft and a for‑profit model.
  • The other side frames it as a simple disagreement where OpenAI chose Microsoft over Musk’s conditional funding.

Musk’s Role, Image, and Competence

• Heated debate over whether Musk “builds things” or is primarily a money/hype person.
• Supporters cite SpaceX’s reusable rockets, Starlink, and Tesla’s EV dominance; some engineers and biographies are invoked to argue he is deeply technically involved.
• Critics point to Twitter/X mismanagement, overpromising on Tesla FSD, and inflammatory public behavior as evidence he’s a poor manager whose wealth distorts his judgment.

Apple–OpenAI Partnership and Grok

• Some speculate Musk’s recent attacks on Apple–OpenAI integration and his lawsuit tactics are driven by jealousy and frustration that his own LLM (Grok) lacks a major platform path.
• Others emphasize Apple’s brand and trust/safety standards as reasons they would avoid Musk-linked AI products.

Legal and Defamation Angles

• Procedurally, Musk dismissed the case without prejudice just before a hearing on dismissal with prejudice, seen by some as a “you can’t fire me, I quit” move.
• A minority argue Apple could consider libel over Musk’s public claims about privacy risks of the ChatGPT integration; others say U.S. free speech protections make such a case implausible.

Static vs dynamic linking and distros

• Big subthread on static linking (Swift static Linux SDK, Rust/Go culture) vs classic distro-managed shared libraries.
• Pro-static side: vendoring dependencies avoids distro-induced version skew, makes builds reproducible, and simplifies distribution (single binary, works across many distros, fewer “dependency hell” issues).
• Pro-dynamic side: distros can patch one shared library to fix many packages at once; rebuilding thousands of statically-linked binaries on a vuln is expensive and often unrealistic.
• Several note that Rust/Go can link dynamically, but tooling and culture strongly push “static everything”; Swift is praised if it allows real choice.

Security, vulnerabilities, and SBOMs

• Critics of static linking highlight past crises (e.g., zlib/xz-like issues): tracking every embedded copy is painful; you often don’t know which version is in a random binary.
• Others argue rebuild storms are acceptable and that binaries should ship an embedded manifest/SBOM of their vendored deps; Go is cited as already doing this.
• Disagreement over whether static linking is a “security nightmare” (opaque dependency graph, slow patch propagation) vs dynamic linking being a risk (uncertain code actually executed at runtime).

Swift on Linux, ABI, and distribution

• Swift binaries on Linux still depend on some system libraries; ABI is explicitly stable only on Apple OSes. On Linux, ABI stability is “unclear” and would likely need distro support.
• New static SDK plus musl-style targets make Swift more competitive with Go for easy container distribution (e.g., Alpine).
• Some see this as aligning with Apple’s broader push: embedded Swift, WASM targets, user-definable platforms.

Language comparisons and ergonomics

• Multiple commenters are enthusiastic about Swift 6: strong type system, no tracing GC (ARC instead), ADTs, protocols (traits), macros, data-race-free concurrency, distributed actors, C++ interop, embedded mode, and upcoming WASM.
• Some describe Swift as “better Go” and “better Rust for app dev,” especially with richer stdlib, async/await, testing, and ergonomics like named arguments and advanced enums.
• Counterpoints: Swift compilation is much slower than Go and often slower than C++ with templates; tooling (SPM configuration, compiler diagnostics, macro setup) has “sharp edges.”
• Others feel the language has become complex and “baby C++”-like, though some argue most developers only need a simple subset.

Ecosystem, server-side, and tooling

• Server-side Swift interest exists (e.g., Vapor), but several say the non-Apple ecosystem is still relatively immature compared to JVM, Go, or Rust.
• Some report painful past experience with server-side Swift (Apple-first design decisions, Linux being second-class), and now prefer Kotlin/Java.
• LSP support for VS Code is described as usable but not yet on par with top-tier ecosystems (missing richer refactors, etc.).

Cross-platform UI and other platforms

• SwiftUI itself is closed-source and Apple-only; third-party SwiftUI-like frameworks (Tokamak, SwiftCrossUI) and wrappers around GTK/Qt/LVGL exist but are seen as not yet mature for serious cross-platform products.
• Embedded Swift and Swift-on-WASM examples are emerging and generate excitement, though they are still early-stage.

Overall reception

• Widely praised as a powerful, convenient, cross‑platform screenshot tool used daily on Linux, Windows, and macOS.
• Users like that it “does what it says,” is quick, and generally stays out of the way until needed.
• Some consider it the best option on Linux; others still prefer alternatives (e.g., Greenshot, ShareX, CleanShot X, Skitch, Shutter) for specific workflows or UX.

Wayland and platform support

• Several report that it now works well on Linux with Wayland.
• Others still experience serious issues (e.g., not working on Ubuntu 22.04 without Xorg, not working under some compositors like Hyprland, feeling less snappy than under X11).
• Works on macOS and Windows, but binding to the Print Screen key is reported as easy on Linux/macOS and unclear or harder on Windows.

Core features and UX

• Strong points: precise region selection, in-place annotations (arrows, text, numbering, highlight, blur, rectangles), quick copy-to-clipboard, and the “pin” feature for floating snippets.
• Numbered callouts are especially appreciated for documentation and support.
• Some find the monochrome, context-dependent toolbar icons unintuitive and dislike that icon layout shifts with selection shape.

Comparisons to other tools

• Compared to ShareX: Flameshot is seen as simpler and less cluttered but missing key features such as built-in video/GIF recording and richer post-capture editing.
• Compared to Greenshot: Flameshot wins on Linux support and in-place annotations; Greenshot is preferred by some for its separate editor, object-level editing, speed, and overall usability.

Workflows and integrations

• Common pattern: bind a shortcut (often Print Screen) to launch Flameshot in selection mode, annotate, then copy or save.
• Several advanced workflows pipe output to scripts: automatic upload to S3 or web hosts, custom image upload APIs, integration with Google Photos or Dropbox, OCR via Tesseract, barcode decoding, and even remote OCR/translation servers.

Limitations and feature requests

• Frequently requested: editable/movable annotation objects, opening existing images, smoothing for freehand drawings, color picker, more upload backends than Imgur, OCR as a native feature, multi-monitor DPI handling fixes, and integrated screen recording (GIF/MP4).
• Some users criticize blur for privacy reasons, preferring cropping or opaque rectangles.

Housing as an Investment (Returns, Costs, and Tax Treatment)

• Many commenters read the paper as confirming that housing has historically matched or beaten equities in real total returns, especially when rents/imputed rents are included.
• Others stress that most casual comparisons ignore full ownership costs: interest, maintenance, renovations, property taxes, and time/effort (“trips to Home Depot”).
• Tax treatment matters: depreciation, 1031 exchanges, mortgage-interest deductibility, and capital-gains exclusions (or absence of tax on primary residences in some countries) all boost effective returns.
• Some argue housing’s strong returns are partly policy-driven (retirement via home equity, political protection of homeowners, zoning constraints).

Population Dynamics and Future Housing Returns

• One camp: global population is still growing (even at a slower rate), so housing demand and land scarcity will keep returns strong for our lifetimes.
• Another camp: fertility rates have fallen almost everywhere; global child counts may already have peaked; population could peak around mid–late 21st century, potentially undermining housing growth.
• Debate remains whether shrinking populations will actually lower prices, given preferences for more space, inheritance dynamics, and rental income.

Housing Costs, Affordability, and “Desirable Land”

• Several posts question how housing can outpace incomes indefinitely, noting a hard ceiling at 100% of household income.
• Responses:
  • Housing quality and size have increased (more space, amenities, codes), so part of the “price increase” is quality.
  • Desirable land (central, coastal, well-serviced) is limited, even if total land is not.
  • Elastic demand: people adjust via roommates, living with parents, smaller units, or moving to less desirable areas.
• Some argue real house prices are usually stable long term, with gains concentrated in bubbles and rents, not prices themselves.

Urban Density, Suburbs, and Zoning

• One view: “density death spiral” — high prices in dense cities drive more density, which further increases land values.
• Counterpoint: construction cost per unit can be lower in dense buildings, but land cost and demand keep total prices high.
• Multiple comments point to zoning and land-use restrictions as key drivers of housing returns and inequality; relaxing zoning is proposed as a major lever, though there’s disagreement on whether any city has truly “built its way out” of high prices.

r vs g, Inequality, and Piketty

• The paper’s finding that returns on capital (especially housing) exceed economic growth is linked to rising inequality and political power of asset owners.
• Some note critiques: if housing is stripped out, capital returns may be closer to or below growth, implying housing policy/zoning is central to the r > g story.

Everyday impacts of better batteries

• High-capacity Li‑ion packs make very loud portable speakers and car-camping power stations common; people note both convenience and nuisance (noise pollution).
• E‑bikes and e‑cargo bikes are seen as “life‑changing” mobility for many trips, with debate over whether they give more or less exercise than non‑electric bikes. Several argue e‑assist increases total riding, especially for older or less-fit riders and hilly areas.

---

Home storage: cost, components, and accessibility

• Users question why residential systems like Powerwall cost ~5× cell prices.
• Explanations: integration (inverter, charger, MPPT, BMS), UL/listing, installation labor, brand/luxury markup, and regulatory overhead.
• DIY and rack-mount LiFePO₄ (LFP) setups are cited at ~€100–300/kWh, but require separate inverters, chargers, safety engineering, and usually licensed electricians to integrate with home panels.
• Some want a “plug‑and‑forget” backup solution; others note that at very small scale this is essentially a UPS, and anything larger needs proper wiring and inspections.

---

Battery safety and fire risk

• LFP is widely praised as much safer than older Li‑ion chemistries; some expect thermal‑runaway‑prone chemistries to fade within a decade.
• Counterpoint: even LFP can off‑gas and be combustible; large stationary packs and EVs still pose nontrivial fire and insurance risks.
• Debate over EV vs gasoline fire statistics: pro‑EV commenters cite much lower fire incidence, skeptics note fleet-age differences, different driving patterns, severity of Li fires, and higher insurance costs.

---

New chemistries: sodium and solid‑state

• Sodium‑ion batteries are already on Chinese marketplaces. Pros cited: cheaper, safer, abundant sodium; cons: heavier and bulkier, steeper discharge curves.
• Seen as especially promising for stationary storage where weight/volume matter less.
• Solid‑state batteries: claims of high safety, very fast charging, long cycle life. Cost and manufacturability remain unclear; early estimates suggest 3–4× Li‑ion initially, with hope economies of scale drive this down.

---

Economics, overcapacity, and the grid

• Commenters stress that the “decade of the battery” is driven more by scale and investment than a single tech breakthrough.
• Reports of massive global overbuild of manufacturing capacity suggest upcoming price drops and possible producer shake‑out.
• Growing distributed storage (EVs, home batteries) is seen as a nascent “virtual power plant,” pre‑distributing large amounts of energy and influencing grid planning.
• Some criticize utilities and regulators (e.g., solar policy changes, high rates), motivating desires to leave or minimize reliance on the grid.

Scope of the Proposal

• Proposal: Ban medical debt from appearing on credit reports.
• Some unclear details in thread: whether all medical-related debt is excluded or only certain categories; commenters speculate but do not know.

Reasons to Support the Ban

• Medical debt is typically involuntary, sudden, and opaque; unlike credit cards or student loans, people rarely “choose” it in a meaningful way.
• U.S. billing is often described as predatory or fraudulent: no upfront prices, post‑facto charges, surprise third‑party bills, and error‑prone collections.
• Patients in emergencies or severe pain cannot realistically negotiate contracts, shop around, or understand intake forms that effectively function as blank checks.
• Cited CFPB data: a large share of people with only medical collections otherwise have clean credit and reliably pay other debts; their medical debt may not be a good predictor of credit risk.
• Many see it as a pragmatic, incremental fix given that broader healthcare reform is politically blocked.

Concerns and Criticisms

• Some argue “debt is debt”: any legal obligation to pay affects ability to service new loans, so lenders should see it.
• Worry that hiding medical debt distorts risk assessment, encourages over‑lending, and may raise interest rates for everyone else.
• Free‑speech angle: restricting reporting of true debt information is characterized by some as limiting commercial speech, though others counter that commercial speech is already regulated (e.g., fraud, privacy laws).

Underlying Healthcare-System Problems

• Multiple commenters stress this treats symptoms, not causes: extremely high U.S. healthcare costs and lack of price transparency.
• Existing “price transparency” rules are seen as toothless; hospitals respond with unusably wide price ranges.
• Suggestions: enforce real upfront pricing with binding quotes, limit or restructure collections, and treat abusive billing as fraud.
• Broader normative view: medical debt “should not exist” and nobody should be financially ruined for needing care.

Related Policy Ideas

• Calls to:
  • Make student loans dischargeable in bankruptcy (and, for some, offer free public college).
  • Move toward single‑payer or other universal systems.
  • Ensure bankruptcy remains available for medical debt, unlike current student debt rules.

Workload, Hours, and On‑Call

• Routine overtime, weekend work, late‑evening interviews, and “we hustle / save the day” narratives are major red flags.
• Required on‑call, frequent 3am pages, or crisis firefighting as normal practice signal poor planning.
• Expectations to attend many after‑hours “social” or “happy hour” events are seen as intrusion on personal life.

Interview Process & Candidate Treatment

• Time‑intensive take‑home projects (4–6+ hours, multiple case studies) are widely criticized as exploitative, especially when uncompensated or used as early screens.
• Some prefer well‑scoped, short (≤1h) take‑homes over LeetCode or live coding; others reject any tests as distrustful or biased toward people without other obligations.
• Negative signals: many short rounds with many teams, rigid templates, “gotcha” trivia, whiteboarding with silent or combative interviewers, cameras off, interviewers distracted or late, or not knowing the role.
• Outsourcing technical screens, automated/chatbot steps, or demanding work on unfamiliar tech despite prior agreements (e.g., specific framework/OS) are seen as disrespectful.
• Lack of time for candidates’ questions, or evasive answers, is a strong red flag.

Culture, Management, and Values

• Phrases like “we’re a family,” “work hard, play hard,” “ninjas/rockstars,” or “mission over money” often indicate unhealthy expectations or underpay.
• Mocking side projects, displaying arrogance or power‑tripping, or signaling intolerance of dissent or alternative solutions are red flags.
• Signs of dysfunction: co‑CEOs with conflicting visions, many management layers, internal communication chaos, nepotism, or leaders threatened by strong candidates.
• Emphasis on face time, open offices, hot‑desking, “frat bro” alcohol culture, dress codes (e.g., suits in casual tech) and lack of remote flexibility can indicate poor fit or outdated thinking.

Technical Practices and Environment

• Lack of tests, fragile release processes, no CI/CD, monolithic databases for many services, very old tech stacks, or “if it ain’t broke we never change” attitudes suggest low engineering maturity.
• Outdated or poor equipment, worn offices, or lax security (e.g., no badges) are read by some as negative; others see them as neutral or even signs of a stable, product‑focused company.

Compensation, PTO, and Career Stage

• Low offers justified by “belief in the mission,” rigid compensation structures, or non‑negotiable boilerplate (especially restrictive clauses) are major warning signs.
• Unlimited PTO is ambiguous: acceptable only when people demonstrably take substantial time off.
• Several posters caution early‑career folks that being this selective may not be realistic; trade‑offs can make sense to build experience.

Nostalgia for the London–Calcutta Route and Overland Era

• Many express fascination with the bus as a symbol of a lost era of casual, long-distance overland travel (hippie trail, adventure buses, desert routes).
• Personal stories: motorcycle/ambulance trips in the 60s–70s, Adelaide–London drives, Kathmandu–London and Johannesburg–London overland tours.
• Several lament not being able to do similar trips now, or missing chances to visit places like Kyiv or Chernobyl before recent crises.

Safety Then vs Now

• One view: the “middle section” countries (Turkey–Iran–Afghanistan–Pakistan region, parts of the Middle East and Africa) are now too unstable or dangerous, making such routes effectively impossible.
• Counterview: globally, travel is statistically safer; many countries (Eastern Europe, much of South/Southeast Asia) and infrastructure have improved. Problems are more about specific bottleneck states (Iran, Pakistan, Afghanistan, some African and Middle Eastern countries).
• Debate over whether perceived danger is due to real deterioration (e.g., Islamist violence, civil wars) or heightened risk sensitivity and media coverage.

Geopolitics and Responsibility

• Some blame local “bad use” of territory; others argue Western colonialism and interventions heavily contributed to regional instability.
• Discussion notes that destabilization and migration also affect Europe in return.
• Concerns about Iran in particular: risk of Westerners being detained, visa unpredictability, and political rather than personal-safety risks.

Current Overland Options and Practicalities

• Descriptions of still-possible routes:
  • Europe–Turkey–Iran–Pakistan–India by train/bus, with Zahedan–Quetta as a harsh, partially unpaved desert segment; Quetta described as very dangerous for Westerners.
  • Alternative trans-Asia rail/bus routes via China, Southeast Asia, and Kazakhstan, with political and legal caveats (e.g., Xinjiang, Myanmar, Russia).
• Visa hurdles (especially Iran) and blocked corridors (Afghanistan, Syria, parts of Africa) make continuous, classic-style London–India or Cape Town–Europe trips very difficult.

Tourism Growth and Cultural Change

• Strong nostalgia for a pre–mass tourism world:
  • Once-quiet attractions like the Acropolis and major museums now heavily crowded.
  • City centers turning into “theme parks” with Airbnbs displacing locals.
  • Perceived cultural homogenization of music, food, language, and urban life; Japan cited as having shifted heavily toward Western norms.
• Others note that many less-famous places (e.g., parts of India, Arunachal Pradesh, rural Asia) still feel untouched and sparsely touristed.

Communication, Technology, and Risk Perception

• Earlier travelers often went incommunicado for weeks; boarding-school and expedition anecdotes highlight how normal delayed communication once was.
• Modern expectations favor constant reachability via cell phones, satellite communicators, and internet research; some see this as increasing safety and enabling better route planning, others as feeding anxiety.
• Debate on whether parents today are overprotective versus realistically responding to changed conditions (e.g., stories of Baja surf trips then vs now).

Modern Long-Distance Buses and Logistics

• Discussion of today’s longest bus routes:
  • Rio de Janeiro–Lima as a very long regular service.
  • A 12,000 km Istanbul–London “bus journey” framed more as a multi-country tour than a practical line.
  • A newly marketed India–London bus project.
• Over-water crossings historically done by ferries or ships; today cars can be shipped as cargo rather than via true passenger car ferries.

Artifacts, Images, and Media Tie-Ins

• Some disappointment that the original article lacks photos; others share links showing the double-decker bus with sleepers, kitchen, and even era-appropriate airline ads.
• Related works mentioned: books on 1960s Switzerland–India road trips, TV series and films about the hippie trail and irregular migration, and historical accounts of desert bus services with colorful, hard-living drivers.

Overview of ARC Prize and Goals

• $1M+ competition centered on the ARC-AGI benchmark: tiny colored-grid puzzles where a system must infer a rule from a few input/output examples and apply it to a new case.
• Intended as an AGI-relevant test of sample‑efficient, on‑the‑fly reasoning rather than large‑scale pattern memorization.
• Main leaderboard runs on Kaggle with limited compute and no internet; a separate unconstrained public leaderboard also exists.

Nature of ARC Tasks: Spatial vs General Intelligence

• Many note that tasks are highly visual/spatial (shapes, containment, symmetry, denoising), raising concern they test human visual priors more than abstract reasoning.
• Others argue nearly all reasoning is ultimately about relationships in space‑time; spatial reasoning is a reasonable core substrate for broader abstraction.
• Comparisons are made to IQ tests and Bongard problems; some see ARC as another narrow domain, not “AGI-complete.”

Comparison to LLMs and Existing AI

• Consensus that direct LLM prompting performs poorly (single‑digit %); even large synthetic finetuning only modestly improves scores.
• Debate over whether LLMs are “expert systems” in disguise, versus a qualitatively different statistical learner.
• Several suggest ARC primarily exposes lack of active, iterative reasoning and working memory in current transformer architectures.

Human Performance and Puzzle Design Issues

• Cited studies show average humans solve ~85%+ of tasks; an easier derivative benchmark still filters out a noticeable fraction of participants.
• Some users find tasks intuitive and quick; others hit ambiguous or seemingly buggy puzzles where multiple answers look valid.
• This fuels criticism that tasks sometimes measure “guessing the test‑setter’s intent” rather than objective correctness.

Prize Money, Incentives, and Data Use

• Mixed views on the $1M prize: some see it as trivial relative to AGI’s stakes; others as mainly advertising and talent‑attraction.
• Concern that the competition crowdsources valuable research cheaply, similar to past industry contests, but many still find the open benchmark valuable.

Broader Debates on AGI and Learning

• Long meta‑discussion on what counts as AGI: human‑level generality vs “things that look intelligent when humans do them.”
• Arguments about human sample‑efficiency (children vs LLMs), the role of evolution as pretraining, and whether intelligence must be grounded in real‑world knowledge.
• Some propose that true progress will require architectures supporting learning at inference time, richer world models, and possibly multi‑agent or human‑in‑the‑loop systems.

Confirmation of death and sources

• Initial skepticism about the Wikipedia change leads to requests for citations.
• Confirmation comes from a university obituary, a major newspaper article, and an early blog post, which Wikipedia then cites.
• Cause of death is reported as a heart condition at home.

Technical and educational contributions

• Widely credited as co-creating modern VLSI design methodology and co‑authoring the seminal textbook “Introduction to VLSI Systems.”
• Her 1978–79 university courses let students design and fabricate their own chips, spawning influential projects like the Geometry Engine and a Scheme microprocessor.
• Commenters explain how the Mead–Conway abstractions transformed hardware from hand‑tuned transistor layouts to standardized cells, libraries, and toolchains, making large designs tractable.
• Earlier at IBM’s ACS project, she authored a 1966 “Dynamic Instruction Scheduling” report, described as foundational for superscalar and out‑of‑order CPUs but largely absent from mainstream textbooks.

Rediscovery, credit, and the “Conway Effect”

• Several note that her early IBM work stayed obscure due to secrecy, project cancellation, and her firing before transition.
• Her own later writing describes how contributions by “outsiders” often disappear or are misattributed, a phenomenon she termed the “Conway Effect.”
• Some express surprise that her ACS work isn’t covered in standard architecture curricula.

Transgender life, IBM firing, and later apology

• Thread repeatedly highlights that IBM fired her in 1968 after she disclosed plans to transition; IBM apologized in 2020.
• Commenters debate whether the apology was too late, but generally see it as meaningful progress.
• Her personal retrospective is recommended as tragic, candid, and ultimately triumphant.

Impact on trans community and representation in tech

• Many trans commenters describe her as a role model whose visibility and resources made their own transitions feel possible.
• Her “success stories” pages and autobiography are remembered as some of the very few online supports in the 1990s–2000s.
• Discussion explores why trans people, especially women, are highly visible in computing: historical online communities, tech’s relative tolerance of “weirdness,” and career restart possibilities.

Wikipedia, deadnames, and policy

• Multiple comments explain why her birth name is absent from Wikipedia: policies on trans biographies, deadnames, and notability under prior names.
• Some argue this is appropriate privacy and safety; others find it encyclopedically unsatisfying but accept HN isn’t the place to litigate the policy.

Politics, human rights, and “apolitical” tech spaces

• One large subthread argues that her story shows how technical work is entangled with social norms and discrimination.
• Many insist trans rights are basic human rights, not “just politics”; others stress that human rights frameworks themselves are politicized and unevenly applied.
• There is extended debate on:
  • Whether everything is inherently political vs. the value of “no politics at work.”
  • How far communities (e.g., language or open‑source projects) should go in explicit inclusivity.
  • The reality and severity of anti‑trans violence, and how rhetoric affects safety.
  • Trans participation in sports and whether biological differences are decisive; commenters cite conflicting interpretations of limited data.

Broader reflections on gender identity

• Long, nuanced exchanges try to explain gender dysphoria in first‑person terms: bodily alienation, not just social roles.
• Several cis readers say these accounts helped them better understand being “born in the wrong body.”
• Others discuss “cis by default” experiences and variation in how strongly people feel about gender at all.

Personal memories and emotional reactions

• Numerous commenters recount studying from her book, taking derivative courses, or using toolchains influenced by her work.
• A few recall meeting her in person at conferences, describing her as curious, kind, and intellectually expansive.
• Many trans readers explicitly thank her for making their own lives and careers more viable.
• The HN black mourning band is discussed; some only now learn its meaning and review prior honorees.

Legacy framing

• Commenters stress that she achieved “great things twice”: first in hardware and architecture, then in education and activism.
• Several wonder how much more she—and others like Turing or other marginalized scientists—might have done without social repression.
• Overall tone is strongly admiring, with occasional critical notes about delayed institutional recognition but near‑universal agreement on the magnitude of her technical and human impact.

Nature of the leaking water and impacts

• Commenters highlight that the erupting water is extremely salty (5–8× seawater) and likely contains oil and hazardous compounds (arsenic, bromide, strontium, mercury, barium, BTEX, etc.).
• This is seen as “salting the earth,” potentially rendering land unusable for agriculture and posing long‑term health risks via food and water contamination.
• Some note that produced/fill water may be even worse than original fracking fluid due to accumulated contaminants.

What’s causing the pressure and leaks?

• One camp argues this is primarily old, improperly capped wells being repressurized by large‑scale saltwater disposal/injection, not fracking itself.
• Others stress that fracking wastewater injection is very likely part of the causal chain and shouldn’t be downplayed.
• Technical explanations emphasize:
  • Difference between temporary “capped” vs permanently “plugged” wells.
  • Produced water being injected back underground, often into old wells or different formations, creating pressure imbalances, potential earthquakes, and blowouts.
• Comparisons to Pennsylvania “frac‑outs” are disputed; some say those were shallow, poorly constructed wells unlike deep Permian wells.

Regulation, liability, and orphan wells

• Strong criticism of “socialized” cleanup: companies profit, then leave taxpayers and landowners with orphan wells and pollution.
• Calls for requiring pre‑paid cleanup funds/bonds; existing orphan‑well schemes (e.g., Alberta) are seen as underfunded or structurally flawed.
• Many note Texas’s lax oversight and budget cuts (EPA and state agencies), plus loose standards for what counts as “fully plugged.”
• Split estate and mineral rights: surface owners often can’t refuse drilling or injection and may have bought land with rights already severed.

Texas Railroad Commission and regulatory capture

• Multiple comments explain why the Railroad Commission regulates oil and gas (historical evolution from rail to pipelines to oil production).
• There is concern about regulatory capture and inadequate ongoing inspection and plugging of old wells.

Environmentalism, messaging, and public understanding

• Some participants argue sensationalist “blame fracking” narratives are technically wrong and counterproductive.
• Others respond that, for the public, the key fact is that hydrocarbon extraction and waste disposal are polluting land and water; precise mechanisms matter mainly for regulators.
• There’s a meta‑debate on balancing technical accuracy, activism, and building political support for stricter regulation.

Broader energy and climate context

• Thread branches into energy independence vs reliance on OPEC, externalities of fossil fuels, and the need to shift to renewables and electrification.
• Nuclear and carbon sequestration are discussed skeptically, with concerns that long‑term underground storage (of CO₂ or waste) will suffer the same leakage problems as old wells.

Overall reception

• Many commenters find the UI clean and the concept useful, especially the keyboard-centric design and large, resizable preview area.
• Some say they will stick with existing tools that already work well for them, mainly due to trust, open‑source preference, or missing features.

Comparison with existing clipboard managers

• Frequently mentioned alternatives: Alfred, Raycast (including its clipboard extension), LaunchBar, Maccy, Flycut, CopyQ, Keyboard Maestro, BetterTouchTool, Paste, and Clipy.
• Several argue Alfred/Raycast/LaunchBar already meet the listed requirements (keyboard control, search, previews, history).
• A recurring differentiator for this app is the ability to resize the history window and see a much larger text preview; users say competing tools often limit preview size.
• Some prefer single‑purpose clipboard tools over multi‑function launchers; others prefer consolidating into a launcher they already use.

Features, limitations, and requested additions

• Currently text‑only; image and file support are explicitly “on the roadmap.”
• No snippets by design; users are pointed to other tools for that.
• Widely requested features:
  • Excluding passwords and specific applications from history.
  • Customizable keyboard shortcuts (including avoiding conflicts with “paste unformatted” and command palettes).
  • Better handling of very large entries (e.g., big Excel ranges) and optional auto‑purging rules.
  • “Quick edit” of a clipboard item before pasting.
  • Opening the window on the active screen (by cursor or focused window).
• Some users report missed clipboard events; at least one reproduction is unclear to the developer.

Privacy, security, and trust

• Strong concern that any clipboard manager approaches a keylogger; multiple commenters say they only trust open‑source or OS‑bundled tools.
• The developer states:
  • No telemetry or network calls except update checks.
  • Local‑only operation is intended.
  • Plans to open‑source the app.
• Discussion includes techniques like blocking outbound connections (e.g., via firewall) and relying on password‑manager clipboard flags or heuristics to exclude secrets.

Implementation and distribution

• App is built as a hybrid: React UI + C++/Cocoa via a custom Electron‑like SDK, leading to large binaries (~245–600 MB).
• Initial download issues due to a cloud storage billing problem were quickly fixed; a GitHub mirror is planned.

Overall reception & positioning

• Many welcome a modern, open‑source CRM and like the UI and ease of setup.
• Some feel marketing as a “Salesforce alternative” is premature; they see Twenty as a promising CRM, not yet a full platform.
• Several say it fills a gap vs older open‑source CRMs that look and feel dated.

Non‑traditional and personal use cases

• Atypical organizations (academic groups, nonprofits, consortia) want contact/relationship tracking not centered on sales funnels.
• Desired capabilities include: longitudinal tracking of careers, research interests, engagement history, and funding discussions.
• Some suggest this is closer to association management software or personal contact managers than classic CRM.
• Personal users with ADHD or relationship‑tracking needs look for calendar‑integrated “personal CRM” tools.

Features, maturity & limitations

• Early users report alpha‑level rough edges and missing basics like granular permissions and role‑based access; anyone can currently alter data models.
• Lack of clear support for custom automation and complex workflows is seen as a major blocker for many business use cases.
• Questions arise about email/calendar provider support beyond Google.

Salesforce comparison, value, and lock‑in

• Multiple comments stress that competing with Salesforce means matching the “last 20%”: compensation, billing, contracts, reporting, etc., plus its vast ecosystem.
• Salesforce is chosen for integration, platform capabilities (Apex, LWC, AppExchange), reliability, and strong admin/rev‑ops community, not just basic CRM.
• Others note Salesforce’s limitations (slow UI, governor limits, complex contracts, lock‑in) and see room for 80% solutions without those constraints.

Licensing & self‑hosting (AGPL)

• AGPL raises concern for companies that might embed proprietary business logic into a self‑hosted instance; some fear “viral” obligations, others argue this is misinterpreted and only applies to modified AGPL code.
• Some note big companies often avoid AGPL entirely to dodge legal complexity.

Extensibility and programmability

• Strong sentiment that a serious Salesforce competitor must offer a hosted programming layer (triggers, scripts, workflows), not just an API.
• Others counter that many smaller customers just need a flexible CRM, not a full low‑code platform, and that enterprise‑grade programmability can come later.

Roll‑your‑own vs buying

• Building a simple, single‑org CRM is considered easy; the difficulty is supporting many orgs, diverse workflows, and integrations.
• Several warn that rolling your own quickly becomes an internal ERP project; off‑the‑shelf CRMs exist largely because of this complexity.

Product changes & scope

• Thread clarifies that the shutdown affects Google’s US peer‑to‑peer (P2P) payments, not tap‑to‑pay at terminals.
• Tap‑to‑pay and card management are handled by Google Wallet (plus OS‑level components) and continue to work.
• GPay had been used for P2P transfers and some online checkout; those P2P features are being retired in the US.
• In India and Singapore, Google Pay remains a much richer, active P2P platform.

Branding and UX confusion

• Many commenters describe long‑running confusion: Android Pay → Google Pay → Google Wallet, plus multiple “Wallet” and “Google Pay” variants over time.
• Some report adding cards via “Google Pay” historically and now being told to use “Wallet,” with overlapping web UIs (wallet.google.com, payments.google.com) and different balances (Google Pay vs Google Play).
• Several see this as emblematic of Google’s product “ADHD” and siloed teams.

Comparisons with Apple & other markets

• Apple’s P2P via Apple Cash has existed for years; the new “tap two iPhones to pay” feature is discussed as just a new initiation channel, not a new system.
• Some predict Google will relaunch P2P under yet another brand once Apple’s marketing proves demand; others argue it may not be worth the cost.
• Multiple comments highlight that other regions (EU SEPA, UK Faster Payments, Canada Interac, India UPI, Australia NPP, Ukraine card‑to‑card) already offer simple, bank‑level P2P, making US dependence on private apps look backward.

Regulation, business model, and interoperability

• P2P is described as low‑margin and support‑heavy, with high regulatory and anti‑fraud overhead. Some speculate Google wants to avoid this scrutiny.
• Users lament app silos (Venmo, CashApp, Zelle, etc.) and the inability to send money across them, attributing it to lock‑in and profit motives rather than technical limits.
• FedNow is mentioned as an emerging US “SEPA‑like” instant payment rail that might eventually reduce the need for private P2P networks.

User trust and Google’s product strategy

• Frequent shutdowns and rebrands (GPay, Fit API, GSuite free tier, Google One changes) are said to erode trust and deter adoption of new Google services, including AI offerings.
• Some call for explicit long‑term support commitments or “beta” labeling for new consumer products.
• Others note that paid enterprise products (Google Cloud, Workspace) do have formal deprecation windows.

Anecdotes and edge cases

• Individual reports include locked funds, large negative balances assumed to be written off, loss of a Google Fi number cascading into loss of a Google account, and migration fatigue leading some to “ungoogle” their lives or favor simpler, bank‑native solutions or even cryptocurrencies.

Context of the Thread

• Original FDA decision (from 2023) to deny a petition against phthalates in food contact materials is resurfacing and framed as under‑noticed at the time.
• Some suspect renewed attention is tied to broader microplastics / endocrine disruptor coverage and a recent related HN thread.

Perceived Health Risks of Phthalates & Microplastics

• Multiple commenters cite recent studies and reviews linking phthalates to:
  • Endocrine disruption, reproductive issues, lower fecundability, semen quality changes.
  • Metabolic effects, oxidative stress, liver function impacts, obesity, type II diabetes.
  • Asthma/allergic disease, immune changes, neurotoxicity, epigenetic changes across generations.
• Others argue evidence in humans at typical exposure levels is still limited, inconsistent, or not yet conclusive.

Trust in the FDA and Regulatory Capture Concerns

• Strong skepticism toward the FDA: accusations of industry capture, corruption, and prioritizing economic interests over precaution.
• Some point to historical delays on harms (lead, asbestos, etc.) and recent drug approvals as evidence the agency is “asleep at the wheel.”

Standards of Evidence and Regulation Philosophy

• Core debate: Should chemicals be allowed until proven harmful, or restricted until adequately proven safe?
• Several argue for a civil “preponderance of evidence” standard and precaution when exposure is ubiquitous and long‑term effects are plausible.
• Others stress you cannot “prove safety,” that over‑caution historically hindered technological progress, and that every restriction has costs.

Everyday Exposure: Flooring, Packaging, and Food

• Long subthread on PVC/vinyl flooring vs carpet vs wood finishes:
  • Concerns about phthalates in PVC, PFAS in stain‑resistant carpets, polyurethane and other coatings.
  • Some prefer vinyl over carpet due to PFAS and dirt; others recommend avoiding large amounts of PVC indoors.
• Discussion of microplastics from packaging, tea bags, Tupperware scraping, and tire dust; disagreement on how significant these routes are.

Alternatives, “Natural” Materials, and Progress

• Mixed views on “natural = safer”:
  • Some favor pre‑industrial materials (glass, wood, cellulose) and see modern synthetics as suspect.
  • Others note many historical materials (lead glazes, kohl, etc.) were clearly harmful and that life expectancy and overall health are much better now.
• Interest in biodegradable or biologically familiar packaging (cellulose, lignin‑based polymers), but recognition of trade‑offs (shelf life, feasibility).

Measurement, Industry Response, and Data Quality

• Mention of VOC/TVOC consumer meters; some find them anxiety‑inducing and noisy rather than actionable.
• A few note evidence that some segments of the food industry are already reducing phthalate use in tubing and materials, though specifics are limited.
• Thread highlights the difficulty of long‑term, population‑wide causal studies once exposure is nearly universal.

Broader Social and Economic Tangents

• Side discussion on insurers potentially using lifestyle and consumption data (including chemical exposures) to price risk, raising fairness and privacy concerns.
• Debate over whether unequal health behaviors should affect risk pooling vs. supporting “social insurance” where the healthy subsidize the less healthy.

Reaction to Microsoft’s Minesweeper App

• Many commenters are shocked or dismayed that an official Microsoft Minesweeper has:
  • Ads, “watch an ad to continue” mechanics, and pay‑to‑win style elements.
  • Large install size (hundreds of MB) compared to classic versions.
• Some see it as “beyond parody” and “cheap/low‑class” behavior for a very rich company.
• Others note this isn’t new; Solitaire and other built‑in games have had similar ad/subscription models for years.

Monetization, Management Incentives, and Branding

• Common explanation: individual teams are driven by KPIs and P&L, so they optimize their own revenue even at brand expense.
• This is framed as rational under corporate/Wall‑St incentives but short‑sighted for Microsoft’s long‑term reputation.
• Several comments describe this as classic “enshittification” and “milking” a captive legacy audience.

Comparisons to Other Platforms and Past Windows

• Earlier Windows (up to 7) are remembered as paid but not ad‑driven, and less abusive to end users.
• Apple and Google are contrasted:
  • Some argue they wouldn’t put banner ads in core utilities (e.g., Notes, Authenticator).
  • Others counter that both already lean heavily on ad and platform revenue and could slide the same way.
• Linux is frequently mentioned as a cleaner alternative, especially GNOME/KDE ecosystems.

Impact on Users and Demographics

• Casual card/minesweeper games are believed to skew toward older or non‑technical users who may:
  • Tolerate or forget subscriptions.
  • Be confused by account/login changes and ever‑shifting UIs.
• Several note that being the “computer person” for family now means dealing with ads, online accounts, and removed features.

Developer Responsibility and Labor Issues

• Debate over whether engineers implementing these mechanics are “complicit” or just following orders under threat of firing.
• Unionization is suggested by some as a response; others point out practical and coordination difficulties.

Alternatives and Retro Options

• Numerous ad‑free replacements are shared:
  • Web‑based Minesweeper and Solitaire clones, Simon Tatham’s puzzle collection, PWAs, and retro Win9x/XP executables.
  • Some report running old Windows games via Wine or in browsers using emulation.

Broader Concerns About Windows

• Many see this as part of wider Windows degradation: OS‑level ads, nagware, and upsells.
• Some worry about long‑term brand damage and loss of trust; others argue users have few practical alternatives.

Historical and political context

• Many connect Chiquita to the United Fruit “banana republic” legacy: coups (e.g., Guatemala 1954), the Banana Massacre, union‑busting, and broader US corporate–CIA interventions in Latin America and beyond.
• Some broaden this to a critique of US foreign policy: Monroe Doctrine, anti‑left militarism in Colombia, Cuba operations, and pattern of backing dictators vs left movements.
• Others note similar allegations about Coca‑Cola, Firestone, Ford, and various multinationals funding or benefiting from paramilitaries or state terror.

Nature of this specific case

• Multiple commenters note this is a civil Alien Tort Statute case brought by Colombian families, resulting in damages, not a government “fine.”
• Earlier, in 2007, Chiquita pled guilty in a US criminal case for paying AUC; DoJ acknowledged the payments were under duress but still illegal, and the company later exited Colombia.

Duress vs complicity

• One camp: paying “protection money” under threat of killing employees and destroying assets makes Chiquita partly a victim; punishing extorted entities seems unjust and analogous to ransomware dilemmas.
• Opposing camp: they could and should have withdrawn sooner rather than continue business while funding a designated terrorist / paramilitary group; paying and staying is not morally neutral.
• Some highlight that “everyone pays protection” in certain regions; others respond that this cannot excuse financing terror or death squads.

Adequacy and meaning of the damages

• Many see ~$38M as trivial relative to Chiquita’s revenue and profits and worry it just becomes a “cost of doing business” rather than deterrence.
• Some call for prison terms or even capital punishment for executives; others note time elapsed, jurisdiction issues, and that a civil suit cannot impose criminal penalties.

Jurisdiction and corporate personhood

• Significant debate over the Alien Tort Statute:
  • Support: one of the few tools for transnational human‑rights accountability; precedent is valuable.
  • Critique: vague “law of nations” standard, extraterritorial overreach, sovereignty concerns, likely Supreme Court target.
• Side discussion on corporate personhood, responsibility “laundering,” and whether corporations should be treated like natural persons in law.

Broader moral questions

• Arguments over whether US corporations should obey US standards globally, even where local norms or law differ (e.g., bribery, cartels).
• Disagreement on collective guilt: are “all Americans” or consumers of cheap bananas/avocados partially responsible, or only state and corporate decision‑makers?