Notas
Índice cronológico de todas as notas do Scholion.
5359 notas · última:
The Fundamental Law Of Software Dependencies
The article argues that canonical source code should include checksums for every dependency, including transitive libraries and the compiler, so builds become reproducible and trust shifts from distributors to hashes.
Opportunistic Refactoring
Refactoring should happen continuously and opportunistically, whenever code needs cleanup, while tests are green. Small changes done now keep the codebase healthier and reduce slower progress later.
LLMs bring new nature of abstraction
The article argues that LLMs change software development as much as the move from assembler to high-level languages, but also add non-determinism, so programming becomes a different kind of interaction with the machine.
“Founder Mode” do Paul Graham: romantização de fundadores ou microgerenciamento disfarçado?
O texto critica a ideia de “Founder Mode” como mitologia centrada no fundador e defende que empresas fortes dependem de colaboração, responsabilidade compartilhada e poder distribuído no time.
Cortes do Pablo Marçal? Pagar micro-influenciadores por posts patrocinados #ad já foi realidade no BR, eis a diferença
O texto compara o Seeding no Twitter da boo-box com a estratégia de cortes de Pablo Marçal. Nos dois casos, pessoas comuns amplificam mensagens, mas a diferença está na transparência e na conformidade ética e legal.
You're getting your last mile wrong
The article argues that software teams should treat production delivery as part of development from the start, so projects stay potentially shippable instead of facing costly last-mile delays and bugs.
The Crucial Difference Between Nice and Kind Leadership
The post argues that nice leadership avoids conflict and protects comfort, while kind leadership tells the truth, makes hard calls, and supports long-term growth.
Exploring the relationship between Tech Lead and Engineering Manager
The article argues that EM and TL should not split accountability. The EM owns technical direction, while the TL extends the EM’s bandwidth by sharing responsibility for specific decisions and projects.
Books I read in September 2024: The Good, The Great and the Meh
The post reviews three September reads: Carrère’s essay collection, Huston’s engineering leadership book, and Morin’s self-help title, ranking them by how useful and engaging they felt.
Books I read in November 2025
The author reviews three November reads, treating reading as a way to broaden perspective, and explains why Dynamic Reteaming works as a reference on team reorganizations while the other two books left stronger emotional marks.
7 Next-Generation Prompt Engineering Techniques
The article reviews seven advanced prompt engineering techniques, explaining how each one helps LLMs produce more accurate, structured, relevant, or verified outputs.
Progress
The article argues that age and politics can pull people toward protectionism and nostalgia, but history still shows progress and new ideas tending to prevail over time.
Ignoring the Wisdom of Crowds
Crowds are accurate when errors cancel out and there is a correct answer, but committee-style voting suppresses innovation by pushing choices toward the bland middle.
About
The page explains why Liquid Projections was created, emphasizing autonomous projectors, dependency control, async support, and small NuGet building blocks for synchronous and asynchronous projectors.
Improve Your Cover Letter by Comparing It to Your Resume
Tailor your cover letter to complement your resume instead of repeating it. Compare both before sending an application, remove overlap, and use the letter for examples that add to the resume.
Strategy testing: avoid the waterfall strategy trap with iterative refinement
The chapter argues that strategy testing should happen before rollout, on a narrow slice, to refine details, track real impact, and avoid a waterfall approach that hides whether the strategy works.
Implementing API Gateways with Ocelot - .NET
The page explains how Ocelot forwards upstream requests to downstream microservices, how configuration.json defines routes, and why multiple API gateways and BFFs can better fit autonomous teams.
5 Ways Incompetent Leaders Dis-Empower Good People
The page argues that control-oriented leaders drain good people, while skillful leaders expand power by including others, giving choice, and focusing on meaningful work and candid feedback.
Web dependencies are broken. Can we fix them?
The web makes basic dependency management depend on bundlers, import maps, or fragile workarounds. The post argues that dependencies should be first-class and mapped more natively.
The New Artificial Intelligentsia
Ruha Benjamin argues that AI evangelists wrap self-interest in humanistic language, reuse eugenic logics, and justify power, inequality, and harm through inevitability, safety, and future-oriented promises.
gRPC: 5 Years Later, Is It Still Worth It?
The author says gRPC remained the right choice for Torq because it preserved compatibility, enforced standards, and shared generated code, while Buf, BSR, ConnectRPC, and Linkerd improved the stack.
Elixir Makes You Make Good Decisions
The author argues that Elixir, Phoenix, Ecto, and the BEAM helped screen.garden make cleaner architecture, easier authorization, faster file sync, and tests that are pleasant to write.
Networked Thought
The page argues that a digital garden should help ideas connect, evolve, and generate new thoughts. It favors low-friction capture, concept linking, and public writing over rigid folders and static archives.
the terminal of the future
The post argues that a better terminal should combine shell integration, persistent sessions, structured I/O, undoable transactions, and a Jupyter-like frontend, then adopt those pieces step by step.
Friction Was the Feature
The essay argues that AI makes words cheap, which floods hiring, admissions, reviews, and other systems with polished but low-value output. The fix is to replace prose with harder-to-fake proof and keep friction where it preserves signal.
Why and How Marten is a Great Document Database
The post shows how Marten stores and loads C# objects as JSON in PostgreSQL, creates missing database objects on demand, and fits the .NET host model while keeping ACID transactions and rich document features.
Thinking in Actors - Part 1
The post argues that web apps are better modeled around real business operations and stateful domain behavior than around CRUD, and presents the Actor Model and virtual actors as a way to handle concurrency, distribution, and fault tolerance.
How to Spot Toxic Feedback: 7 Signs That the Writing Advice You’re Getting May Do More Harm Than Good
Susan DeFreitas lists seven signs that writing feedback becomes toxic when it misses the writer’s intent, projects personal issues, turns personal, uses an unprofessional tone, relies on formulas, skips real problem-solving, or confuses diagnosis with fixes.
Night Shift Agentic Workflow
Jamon Holmgren describes a day-shift/night-shift setup where he prepares specs during the day and lets AI agents work overnight with strict tests, docs, and review loops.
All I Need to Know About Engineering Leadership I Learned From Leave No Trace
A articulation of Leave No Trace with engineering leadership becomes seven principles applied to software: plan, use durable technology, clean up technical debt, give back improvements, and treat security as care with broad impacts.
An introduction to thinking about risk
The series defines risk as the combination of likelihood and impact, shows when to measure it numerically or with simple matrices, and prepares the discussion on mitigation.
How to Create a Learning Culture: 9 Leadership Strategies for Defining Workplace Success
A page outlines nine leadership patterns for building a learning culture: celebrate learning moments, support experimentation, make time for learning, and use OKRs, vulnerability, MVPs, audits, pain, and intent to reinforce growth.
A Culture of Good Decision-Making (Part 1)
The article argues that better decisions come from separating important from minor choices, decentralizing authority, defining ownership, and sharing context so teams can decide faster with less friction.
Life as a physical process
The text proposes treating life as a physical process that generates systems statistically unlikely to assemble and persist, using Assembly Theory to rethink life, aliveness, and even AI.
Software Is Dead — Long Live Software
A recent software sell-off is treated as evidence that SaaS is dying, but the piece argues that this reading is too broad. It says that AI is still software, the cloud and automation expand markets, and the defense comes from workflow, data, and trust.
Personal Finance for Engineers 🏦
The article argues that personal finance matters for engineers because they often earn well, face career volatility, and can benefit a lot from starting early. It ties money habits to long-term outcomes like savings, retirement, and flexibility in choosing jobs.
Measuring personal growth
The post proposes three ways to think about personal growth: becoming a different person every 3–6 years, solving major life problems faster, and choosing actions that preserve or expand future options.
How to handle external dependencies of your domain in a command-event architecture
The article compares five ways to handle external dependencies in a command-event architecture domain, from classic injection to the reader monad, and concludes in favor of internal commands with explicit dependencies.
A crash course in memory management
Explains memory as fixed-size boxes with addresses, contrasts JavaScript’s automatic garbage collection with manual allocation and freeing in C, and links that tradeoff to performance and bugs.
The myth of software time estimations
The article argues that software time estimates are unreliable because development is inherently unpredictable, and says teams should avoid deadlines in favor of difficulty estimates and Kanban.