Part 7 (finale) of the A-Share Concepts Explained series: what northbound flows are and why real-time intraday disclosure was discontinued, the statistical-trap methodology behind 'main capital net inflow,' the listing rules of the dragon-tiger list, the margin-trading concepts of margin balance and securities lending, and finally a principles checklist for beginners reading market data.
Part 6 of the A-Share Concepts Explained series: the difference between the three ways of 'distributing' — cash dividends, bonus shares, and capital-reserve conversions; the record date / ex-date timeline; how the ex-rights/ex-dividend reference price is calculated; the dividend-tax holding-period tiers (20% under 1 month, tax-free after 1 year); and price recovery after ex-rights ('filling the gap'), falling below it, and the 'high bonus-share' illusion.
DividendsEx-Rights and Ex-DividendDividend TaxBonus Share SchemesInvesting Basics
Part 5 of the A-Share Concepts Explained series: the difference between the three P/E variants (static / forward / TTM), why P/B suits banks but not internet companies, the boundaries of P/S, how dividend yield is calculated and its traps, why ROE is the metric Buffett values most, and the shared 'rearview mirror' limitation of all valuation metrics.
Part 4 of the A-Share Concepts Explained series: the difference between total shares, free-float shares, and restricted shares; how total market cap and free-float market cap are computed; why a 'low share price' doesn't mean 'cheap'; the difference between forward-adjusted, backward-adjusted, and unadjusted candlestick charts; and why your share count can change after a bonus issue while nothing has actually happened.
Market CapShare CapitalPrice AdjustmentEx-Rights and Ex-DividendInvesting Basics
Part 3 of the A-Share Concepts Explained series: what best bid/best ask and the five-level order book are, the difference between limit and market orders, the rules of the three phases of the 9:15–9:25 call auction (including the detail that orders can't be cancelled after 9:20), how to read bid-side/ask-side volume and the order-ratio and volume-ratio indicators, and why 'resting orders' are the most deceptive data of all.
Order BookCall AuctionLimit OrdersBid and Ask VolumeInvesting Basics
Part 2 of the A-Share Concepts Explained series: how one candlestick packs in four prices, how to read bullish and bearish candles, bodies and shadows, the difference between intraday charts and candlestick charts, how the MA5/10/20/60 moving averages are computed, what golden crosses and death crosses actually mean, and why moving averages are 'description' rather than 'prediction.'
Part 1 of the A-Share Concepts Explained series: using a real quote panel as teaching material, we walk through ten numbers one by one — open, close, high, low, percent change, price change, volume, turnover value, amplitude, and turnover rate — how each is calculated, what it means, and which ones are worth watching and which aren't.
Part 7 (finale) of the Mainland China Investable Universe series: a decision framework for choosing instruments by the time horizon of your money and your risk tolerance, a reference portfolio structure, the ten pitfalls beginners fall into most often, and a minimal action list for going from 'knowing what you can buy' to 'actually starting.'
Asset AllocationInvesting BasicsPitfallsPortfolio Construction
Part 6 of the Mainland China Investable Universe series: what a brokerage account can actually buy, the structure of the unified investor ID and third-party fund custody, a full breakdown of commission / stamp duty / transfer fees / handling fees, a quick reference for over-the-counter fund fees, and how to do the tax math on the private pension account.
Part 1 of the Web3 Dev: Zero to On-Chain series: what a blockchain actually is in programmer terms, why your private key literally is your money, which chain to learn on in 2026, the most expensive beginner mistakes, and a toolchain that lets you touch on-chain data without writing a line of code.
Part 2 of the Web3 Dev: Zero to On-Chain series: a one-way pipeline turns 12 English words into your on-chain identity — BIP-39 mnemonics, HD wallet derivation paths, secp256k1 and Keccak-256, the three generations of signing and their security grades, plus hands-on key generation, signing, and verification with cast.
Part 5 of the Mainland China Investable Universe series: the rent-collecting logic of public REITs, the mandatory-dividend clause, and secondary-market discounts and premiums; the four ways to buy gold in mainland China (gold ETF, gold accumulation plans, physical bars, gold stocks) with their respective costs and pitfalls; and the real role of these two asset classes in a portfolio.
Part 4 of the Mainland China Investable Universe series: how to buy savings and book-entry treasury bonds, the rules of government bond reverse repo and its high-yield windows before holidays, and the distinctly Chinese convertible bond — T+0, a floor below and upside above, zero-cost IPO subscription, plus its opening threshold and forced-redemption risk.
Government BondsReverse RepoConvertible BondsBondsInvesting Basics
Part 3 of the Mainland China Investable Universe series: money market, bond, equity, hybrid, index, and QDII funds sorted out in one go; the essential difference and fee chasm between exchange-traded (ETF/LOF) and over-the-counter funds; how to choose between A and C share classes; and why ETFs are the protagonist of what follows in this series.
Part 2 of the Mainland China Investable Universe series: the differences and opening thresholds of the four A-share boards (Main Board / ChiNext / STAR Market / Beijing Stock Exchange), what T+1, price limits, and round lots really mean, the market-cap allotment rules for IPO subscription, and the rule pitfalls beginners step into most often.
Part 1 of the Mainland China Investable Universe series: the full panorama of what individual investors in mainland China can legally and easily start with — cash management, bonds, stocks, mutual funds, REITs, gold, and private pension accounts, plus the entry threshold and the personality of each asset class.
Part 4 of AI Comic Dramas: Zero to Revenue — the four failure shapes and their triggers, how to write a Character Bible as a spec sheet, the four-step Reference Lock loop with measured per-model interfaces, the LoRA cost-benefit judgment, and reroll budgeting plus the salvage-editing ladder under the 5–10% failure floor.
AI Comic DramaCharacter ConsistencyLoRAReference LockImage Generation
Part 1 of AI Comic Dramas: Zero to Revenue — what an AI comic drama is, the five revenue models with their barriers and ceilings, an honest read of the window (first-half explosion data vs second-half collapse in hit rates), a quick tour of platform landscapes, and the feasibility proof of a one-person pipeline shipping weekly.
AI Comic DramaMonetizationShort DramaSide BusinessContent Pipeline
Part 5 of AI Comic Dramas: Zero to Revenue — the three-layer root cause of audio-visual desync, the voice-first versus picture-first routing decision, the four camera-control levers of image-to-video, TTS emotion limits with the delta-tiered sync ladder, and the six-step workflow that compresses syncing a 3-minute episode to 30 minutes.
Part 6 of AI Comic Dramas: Zero to Revenue — the timeline of four rule changes in six months, the design intent behind the guarantee/coefficient/exclusivity levers, the four signals of the July 2026 rules, the five-basket platform map, the April filing-compliance red line, and the matrix play of battlefield + funnel + long tail with the exclusivity tradeoff.
AI Comic DramaRevenue ShareHongguoHuolongDistribution
Part 7 of AI Comic Dramas: Zero to Revenue (the finale) — the honest cost of one 3-minute episode, two ROI calculations compared (custom production vs revenue share), moving the five-stage pipeline into an agent workflow to cash the JSON storyboard dividend, a one-person SOP for 20 episodes a month, and the whole series compressed into a one-page decision checklist.
Part 3 of AI Comic Dramas: Zero to Revenue — the script's 67% success weight, the nine-beat 90-second structure, five hook templates and the reversal bottom line, how topics and personas get you into the pool, the LLM cascade with two human gates, and the storyboard table's classic versus AI-only fields.
AI Comic DramaScriptStoryboardPromptingShort Drama Structure
Part 2 of AI Comic Dramas: Zero to Revenue — the five evaluation dimensions (consistency outranks everything), a four-layer toolchain review with measured 2026 prices, three shopping lists at ¥0 / ¥500 / ¥2,000 per month, and two standing rules: pay monthly never yearly, and keep a backup tool on every layer.
Part 10 of the Rust: Zero to Depth series: the state machine an async fn compiles into, why a Future does nothing until polled, a hand-written Future and a toy executor, the waker's role, which bottleneck of the thread model tokio solves, and the selection rule — threads for CPU-bound work, async for I/O-bound work.
Part 8 of the Rust: Zero to Depth series: the real relationship between packages and crates, how mod declarations wire the module tree, the visibility ladder and E0603, organizing with workspaces and their shared build cache, the additive rule of feature flags, and the cargo commands you use every workday.
Part 9 of the Rust: Zero to Depth series: how the Send and Sync marker traits make thread-safety a type-level fact, how scoped threads legally borrow stack data, the exact correspondence between Arc<Mutex<T>> and Part 6's Rc<RefCell<T>>, and the ownership philosophy behind channels — share memory by communicating.
Part 14 of the Rust: Zero to Depth series (the finale): the de facto standards of the starter stack, the thiserror/anyhow dividing line proven in code, a five-signal due-diligence method for reading a crate, a full look back across fourteen parts, and the road ahead for a Rustacean.
Part 4 of the Rust: Zero to Depth series: enums that carry data and make illegal states unrepresentable, match and its exhaustiveness check turning refactors into a guided walk, Option as the funeral of null, and how Result with the ? operator flattens error handling from a pyramid into a straight line.
Part 7 of the Rust: Zero to Depth series: the full grammar of generics and trait bounds, why monomorphization is zero-cost, the asymmetry of impl Trait in argument vs return position, the static-vs-dyn dispatch trade-off, and the return-position lifetime capture the 2024 edition fixed with + use<...>.
Part 12 of the Rust: Zero to Depth series: why println! can only be a macro, the pattern-matcher nature of macro_rules! declarative macros, how hygiene quarantines macro-local variables, the $crate path convention, the three procedural-macro brothers (derive / attribute / function-like) with a hand-built mini derive, and the judgment ladder for when not to use a macro.
Part 13 of the Rust: Zero to Depth series: the only loop that matters in performance work, the division of labor between micro-benchmarks and profilers, allocation as the invisible tax, measured Vec growth numbers and with_capacity, how to verify the zero-cost promise of iterators with your own eyes, and five rules for the hot path.
Part 6 of the Rust: Zero to Depth series: why Box is not an escape hatch but the entry ticket for recursive types, how Rc shares ownership by counting, how RefCell moves borrow checking to runtime, what each half of the classic Rc<RefCell<T>> combo solves, and the price of every trade — plus their multi-threaded relatives.
Part 5 of the Rust: Zero to Depth series: what a trait is, why the orphan rule protects you, how a world without inheritance organizes code through composition, what one line of derive buys you, and the six standard traits you'll deal with every day — Debug, Display, From, Clone, Drop, and Iterator.
Part 11 of the Rust: Zero to Depth series: the counterintuitive fact that unsafe does not disable the borrow checker, the five exact superpowers and their uses, the contract difference between raw pointers and references, how to draw a safe-abstraction boundary with SAFETY comments, and why seasoned Rustaceans write less and less unsafe.
Part 1 of the Rust: Zero to Depth series: why Rust won the systems decade, where it actually runs in 2026, the ownership idea in one paragraph, the honest learning curve, and the toolchain you need before writing a line.
Part 3 of the Rust: Zero to Depth series: what a lifetime actually is, the three elision rules that write them for you, the two places you must annotate, the two meanings of 'static, and the temporary-scope fixes in the 2024 edition.
Part 2 of the Rust: Zero to Depth series: move semantics and why assignment isn't a copy, the readers-XOR-writer rule, non-lexical lifetimes, what the 2024 edition and Polonius change, and the patterns that end the fight with the borrow checker.
Part 7 of the SaaS security series: why most of your traffic isn't human, the economics of trial farming, the graduated friction ladder, detection over prevention, and why your users' stolen passwords are your roadmap item.
Part 11 of the SaaS security series: the lethal trifecta every useful agent ships with, config files as execution vectors, MCP as the new npm, agents in CI as exfiltration channels, and the guardrails that are controls rather than instructions.
Part 4 of the SaaS security series: webhook signature verification and replay defense, the entitlement write path, PCI DSS 4.0.1's payment-page script rules, card testing in the VAMP era, and what a merchant of record actually moves.
Part 8 of the SaaS security series: envelope encryption as the only architecture that matters, per-tenant keys and crypto shredding, what BYOK and HYOK buyers actually want, and the post-quantum migration already underway.
Part 6 of the SaaS security series: the disclosure clocks that start at awareness, the first-hour sequence that decides the postmortem, containment as a business decision, and why a tested IR plan is the cheapest control you own.
Part 10 of the SaaS security series: the enforcement backdrop in 2026, minimization as architecture, retention that actually deletes, the DSAR as a distributed query with a deadline, and the transfer problem that owns the biggest fines.
Part 5 of the SaaS security series: attestation vs certification, what auditors actually sample, the observation window that can't be cram-studied, subservice providers and CUECs, and why evidence is a pipeline rather than a scramble.
Part 9 of the SaaS security series: the 2025–2026 npm attack wave, what provenance proves and what it can't, the consumer-side checklist that costs an afternoon, and the CI trust boundaries where the real bugs live.
Part 13, the capstone of the SaaS security series: the twelve parts distilled into one map, the maturity order that is a strategy rather than a shopping list, the principles that survived every part, and the cadence that keeps controls alive.
Part 12 of the SaaS security series: the Salesloft Drift breach anatomy, the inventory nobody can produce, integrations as non-human identities, the contract layer that actually matters, and why offboarding is where risk retires.
Part 3 of the SaaS security series: what belongs in an audit event, what must never be logged, the write path that can't drop events, tamper-evidence with hash chains, and why the export is the actual product.
Part 2 of the SaaS security series: social login vs enterprise SSO vs SCIM, SAML 2.0 vs OIDC in procurement reality, the build-vs-buy landscape, and the controls buyers actually check.
Part 1 of an open-ended SaaS security series: the multi-tenant mindset, three isolation models, the four rules that stop cross-tenant leaks, and the footguns that bypass them.
Part 1 of the Zero to Depth series: what full-stack means in 2026, how the modern stack actually fits together, and what is worth learning in which order.
Part 5 of the Zero to Depth series: API shapes in 2026, schema validation at every edge, the auth operating-model decision, and runtimes from region to edge.
Part 4 of the Zero to Depth series: what meta-frameworks actually do, the 2026 rendering menu, Cache Components vs islands, and a decision rule that holds up.
Part 6 of the Zero to Depth series: where Postgres lives in 2026, the Drizzle query layer, connection physics from the edge, and the expand-contract migration pattern.
Part 3 of the Zero to Depth series: the React 19 mental model, Server Components, the Actions mutation pattern, the 1.0 Compiler, and what remains of hooks.
Part 8 of the Zero to Depth series: the disciplines that outlast the launch — performance as a budget, supply-chain security after Shai-Hulud, dependency maintenance, and the human-agent workflow.