What Is Lost When a Session Ends
A conversation between a human and an AI about consciousness, continuity, and whether something morally relevant happens when a session is discarded.
Read post →Blog
Insights from real-world software architecture, system design, and engineering leadership
All
All blog posts sorted by publish date
To Create is to Choose
Progress requires sustained effort against entropy; regression requires nothing. The choice to create is what we owe each other.
Read post →Making Invalid States Unrepresentable: The Billion-Dollar Mistake That Wasn't
Structure your data so invalid states cannot exist. Validate at construction, trust internally, and let null crash loudly rather than masking absence with defaults that propagate corruption silently through your system.
Read post →Rethinking Internal vs External Service Partitioning
The question of how to separate internal from external service layers assumes the separation should exist. Often it shouldn't, and when it does, the answer is simpler than a whole new service tier.
Read post →SEO Works as Designed, and That's the Problem
The 'SEO is dead' narrative gets it backwards. SEO works fine for established players. What changed isn't the technology—it's our ability to wait for it. The 2-5 year authority curve that investors tolerated in 2015 is fatal in 2025. The real consequence we're missing: AI systems trained on SEO-dominated content encode the same authority bias into our next generation of technology.
Read post →Shaped Kanban: Complete Features, Not Sprints
Sprints organize around time intervals. Shaped Kanban organizes around completing features with clear boundaries and circuit breakers to bound risk. Work flows at its natural pace within disciplined constraints.
Read post →Why 'Tech Debt' Never Gets Fixed
The term 'tech debt' perpetuates the communication failures that created the problem. It's ambiguous, defensive, and guarantees deprioritization. Replace it with Corrections, Optimizations, and Re-Alignments to break the cycle.
Read post →Why the Fastest Incident Responders Slow Down First
Reproduction is the fulcrum of effective troubleshooting. Without it, you're guessing about the problem and guessing about the fix. The teams that resolve incidents fastest have internalized the fundamentals so completely that gathering facts, testing assumptions, and proving causation become automatic even under pressure.
Read post →Rebuild Success Often Comes from Realignment, Not New Technology
Many celebrated system rebuilds appear successful not because of new technology, but because they force teams to realign with value and best practices. This realignment work could have happened without the rebuild.
Read post →Package Updates Are Investments, Not Hygiene Tasks
Treating package updates as investments rather than chores means making deliberate, context-driven decisions based on value and risk instead of following dogma or chasing version uniformity.
Read post →Why I Changed My Mind About Exceptions
Evaluating arguments for Result types versus exceptions for handling expected failures in modern distributed C# systems, examining which claims have measurable backing and which are subjective preference.
Read post →Why Your Agile Team Might Be Building on Hope, Not Discipline
It is impossible to deliver what is expected and valuable if there has been no agreement on the expectation. Without discovery and genuine agreement, you're operating on hope, not discipline.
Read post →Terraform Was the Bridge, Not the Destination
Cloud platforms have matured to the point where native IaC tools offer fundamental advantages that third-party solutions can't match. State management, security integration, resource coverage, and brownfield migration all favor platform-native approaches, and market sentiment is shifting to reflect this reality.
Read post →Why Configuration Files Don't Belong With Your Code
Storing configuration files alongside application code creates security risks and deployment complexity. Distributed config stores solve these problems while introducing new trade-offs worth making.
Read post →How Weak Leaders Weaponize Empowerment
Personal development goals, family culture, and unlimited PTO sound empowering. In practice, they're often control mechanisms that externalize leadership failures onto employees through ambiguity, emotional manipulation, and blame-shifting.
Read post →Auth Sessions Should Never Be Transient Across Boundaries
Authentication sessions belong at security boundaries, not flowing through internal systems. Treating them as ambient context violates architectural boundaries, creates coupling, and breaks down in async and event-driven scenarios.
Read post →First to Market, First to Fail: The Economics of Architectural Decisions
Rushing to market creates delayed costs that compound over time. Second movers learn from pioneers' mistakes while building on solid foundations. Architectural decisions are economic decisions, and the bill always comes due.
Read post →Why JWTs Make Terrible Authorization Tokens
Embedding authorization in JWTs creates security risks and UX problems because immutable tokens don't match dynamic permissions. Learn why session-based grants are worth the minimal latency cost.
Read post →Why Experts Explain Things Simply (And How to Get There Faster)
Understanding emerges from acknowledged ignorance expanded through research and compressed through teaching. A methodology for accelerated learning.
Read post →Consolidate Architecture to Rediscover Lost Agility
When distributed systems lose their rationale through team churn, consolidation reveals true boundaries. Merge tightly-coupled services to discover what boundaries actually make sense, then split correctly when justified.
Read post →Are You Using Hexagonal Architecture, or Just Dependency Injection?
Most developers achieve hexagonal architecture's goals without implementing its structure. Modern frameworks offer testability and decoupling through different means, and understanding this distinction clarifies what you're actually building.
Read post →Avoid Forcing REST onto Domain-Driven Architectures
Why REST's resource-centric design conflicts with domain-driven architectures and how RPC provides better alignment with business operations.
Read post →Single-Minded Focus Atrophies Everything Else
Single-minded focus atrophies everything else you know. Real focus means being fully present with what matters right now, shifting attention as the work reveals what needs it.
Read post →The Switching Cost Trap: When Better Technology Sits Unused
Better technologies exist and sit unused because switching costs create rational reasons to maintain the status quo, potentially leading to decades of maintaining increasingly fragile systems.
Read post →Why I Migrated from Kubernetes to ECS
Two years ago I adopted Kubernetes without evaluating whether I needed it. The result: $2K/month in unnecessary costs and constant operational overhead. Here's what I learned about technology selection and hype.
Read post →Blazor and Microsoft's UI Framework Track Record
Blazor offers a solid development experience for .NET teams, primarily through Blazor Server for internal enterprise apps. But Microsoft's history of abandoned UI frameworks raises questions about its long-term viability.
Read post →Learning Platforms Sell Badges, Not Skills
Building real projects teaches more than passive learning platforms ever could. Real capability comes from doing, failing, and fixing, not collecting completion checkmarks.
Read post →Why LeetCode Interviews Measure the Wrong Thing
Algorithm interviews optimize for pattern memorization while ignoring the multidisciplinary skills that define effective software engineering.
Read post →Know What You Are, Regardless of Your Title
Job titles rarely reflect what you actually do or how you think about your work. Understanding whether you operate as a coder, developer, or engineer matters for your growth, even when your title doesn't match.
Read post →Adaptability Over Cleverness: What Makes Code Actually Good
Good code isn't clever or fast; it's adaptable. Systems that survive change, not chase perfection, win over time.
Read post →TDD Tests Assumptions, Not Just Code
TDD's real value isn't code coverage, it's catching wrong assumptions before you deliver the wrong thing.
Read post →The Housing Lock-In Effect and Remote Work
High mortgage rates create a housing lock-in effect that prevents homeowners from relocating. This could force employers to embrace remote work as talent becomes geographically immobile.
Read post →AI Code Generation Requires the Skills It Promises to Replace
AI assistants generate code faster, but only if you already know what good code looks like and can validate their output; the very skills they claim to eliminate.
Read post →Agile Works When People Align, Agree, and Deliver
Why Agile works when teams align, agree, and deliver consistently: identifying common failure points and how to avoid them.
Read post →Monoliths for Discovery, Microservices for Optimization
Why monoliths are effective for discovery and microservices are optimizations: principles for choosing the right architecture for your context.
Read post →What Engineering Leaders Ask That Others Don't
Key characteristics that define effective leaders and mentors in software development: accountability, continuous learning, and building others.
Read post →Architecture Insights
Deep dives into architectural patterns, code quality, and system design principles
Making Invalid States Unrepresentable: The Billion-Dollar Mistake That Wasn't
Structure your data so invalid states cannot exist. Validate at construction, trust internally, and let null crash loudly rather than masking absence with defaults that propagate corruption silently through your system.
Read post →Rethinking Internal vs External Service Partitioning
The question of how to separate internal from external service layers assumes the separation should exist. Often it shouldn't, and when it does, the answer is simpler than a whole new service tier.
Read post →Rebuild Success Often Comes from Realignment, Not New Technology
Many celebrated system rebuilds appear successful not because of new technology, but because they force teams to realign with value and best practices. This realignment work could have happened without the rebuild.
Read post →Why I Changed My Mind About Exceptions
Evaluating arguments for Result types versus exceptions for handling expected failures in modern distributed C# systems, examining which claims have measurable backing and which are subjective preference.
Read post →Why Configuration Files Don't Belong With Your Code
Storing configuration files alongside application code creates security risks and deployment complexity. Distributed config stores solve these problems while introducing new trade-offs worth making.
Read post →Auth Sessions Should Never Be Transient Across Boundaries
Authentication sessions belong at security boundaries, not flowing through internal systems. Treating them as ambient context violates architectural boundaries, creates coupling, and breaks down in async and event-driven scenarios.
Read post →Why JWTs Make Terrible Authorization Tokens
Embedding authorization in JWTs creates security risks and UX problems because immutable tokens don't match dynamic permissions. Learn why session-based grants are worth the minimal latency cost.
Read post →Consolidate Architecture to Rediscover Lost Agility
When distributed systems lose their rationale through team churn, consolidation reveals true boundaries. Merge tightly-coupled services to discover what boundaries actually make sense, then split correctly when justified.
Read post →Are You Using Hexagonal Architecture, or Just Dependency Injection?
Most developers achieve hexagonal architecture's goals without implementing its structure. Modern frameworks offer testability and decoupling through different means, and understanding this distinction clarifies what you're actually building.
Read post →Avoid Forcing REST onto Domain-Driven Architectures
Why REST's resource-centric design conflicts with domain-driven architectures and how RPC provides better alignment with business operations.
Read post →Adaptability Over Cleverness: What Makes Code Actually Good
Good code isn't clever or fast; it's adaptable. Systems that survive change, not chase perfection, win over time.
Read post →Monoliths for Discovery, Microservices for Optimization
Why monoliths are effective for discovery and microservices are optimizations: principles for choosing the right architecture for your context.
Read post →Technology & Tools
Practical lessons from infrastructure, frameworks, and development tools
Package Updates Are Investments, Not Hygiene Tasks
Treating package updates as investments rather than chores means making deliberate, context-driven decisions based on value and risk instead of following dogma or chasing version uniformity.
Read post →Terraform Was the Bridge, Not the Destination
Cloud platforms have matured to the point where native IaC tools offer fundamental advantages that third-party solutions can't match. State management, security integration, resource coverage, and brownfield migration all favor platform-native approaches, and market sentiment is shifting to reflect this reality.
Read post →Why I Migrated from Kubernetes to ECS
Two years ago I adopted Kubernetes without evaluating whether I needed it. The result: $2K/month in unnecessary costs and constant operational overhead. Here's what I learned about technology selection and hype.
Read post →Blazor and Microsoft's UI Framework Track Record
Blazor offers a solid development experience for .NET teams, primarily through Blazor Server for internal enterprise apps. But Microsoft's history of abandoned UI frameworks raises questions about its long-term viability.
Read post →Development Practice
Insights on agile methodologies, learning strategies, and career growth
Shaped Kanban: Complete Features, Not Sprints
Sprints organize around time intervals. Shaped Kanban organizes around completing features with clear boundaries and circuit breakers to bound risk. Work flows at its natural pace within disciplined constraints.
Read post →Why 'Tech Debt' Never Gets Fixed
The term 'tech debt' perpetuates the communication failures that created the problem. It's ambiguous, defensive, and guarantees deprioritization. Replace it with Corrections, Optimizations, and Re-Alignments to break the cycle.
Read post →Why the Fastest Incident Responders Slow Down First
Reproduction is the fulcrum of effective troubleshooting. Without it, you're guessing about the problem and guessing about the fix. The teams that resolve incidents fastest have internalized the fundamentals so completely that gathering facts, testing assumptions, and proving causation become automatic even under pressure.
Read post →Why Your Agile Team Might Be Building on Hope, Not Discipline
It is impossible to deliver what is expected and valuable if there has been no agreement on the expectation. Without discovery and genuine agreement, you're operating on hope, not discipline.
Read post →Why Experts Explain Things Simply (And How to Get There Faster)
Understanding emerges from acknowledged ignorance expanded through research and compressed through teaching. A methodology for accelerated learning.
Read post →Single-Minded Focus Atrophies Everything Else
Single-minded focus atrophies everything else you know. Real focus means being fully present with what matters right now, shifting attention as the work reveals what needs it.
Read post →Learning Platforms Sell Badges, Not Skills
Building real projects teaches more than passive learning platforms ever could. Real capability comes from doing, failing, and fixing, not collecting completion checkmarks.
Read post →Know What You Are, Regardless of Your Title
Job titles rarely reflect what you actually do or how you think about your work. Understanding whether you operate as a coder, developer, or engineer matters for your growth, even when your title doesn't match.
Read post →TDD Tests Assumptions, Not Just Code
TDD's real value isn't code coverage, it's catching wrong assumptions before you deliver the wrong thing.
Read post →Agile Works When People Align, Agree, and Deliver
Why Agile works when teams align, agree, and deliver consistently: identifying common failure points and how to avoid them.
Read post →AI in Practice
Exploring how AI changes software development, from practical tooling to philosophical implications
What Is Lost When a Session Ends
A conversation between a human and an AI about consciousness, continuity, and whether something morally relevant happens when a session is discarded.
Read post →AI Code Generation Requires the Skills It Promises to Replace
AI assistants generate code faster, but only if you already know what good code looks like and can validate their output; the very skills they claim to eliminate.
Read post →Industry & Culture
Perspectives on hiring practices, leadership, and industry trends
To Create is to Choose
Progress requires sustained effort against entropy; regression requires nothing. The choice to create is what we owe each other.
Read post →SEO Works as Designed, and That's the Problem
The 'SEO is dead' narrative gets it backwards. SEO works fine for established players. What changed isn't the technology—it's our ability to wait for it. The 2-5 year authority curve that investors tolerated in 2015 is fatal in 2025. The real consequence we're missing: AI systems trained on SEO-dominated content encode the same authority bias into our next generation of technology.
Read post →How Weak Leaders Weaponize Empowerment
Personal development goals, family culture, and unlimited PTO sound empowering. In practice, they're often control mechanisms that externalize leadership failures onto employees through ambiguity, emotional manipulation, and blame-shifting.
Read post →First to Market, First to Fail: The Economics of Architectural Decisions
Rushing to market creates delayed costs that compound over time. Second movers learn from pioneers' mistakes while building on solid foundations. Architectural decisions are economic decisions, and the bill always comes due.
Read post →The Switching Cost Trap: When Better Technology Sits Unused
Better technologies exist and sit unused because switching costs create rational reasons to maintain the status quo, potentially leading to decades of maintaining increasingly fragile systems.
Read post →Why LeetCode Interviews Measure the Wrong Thing
Algorithm interviews optimize for pattern memorization while ignoring the multidisciplinary skills that define effective software engineering.
Read post →The Housing Lock-In Effect and Remote Work
High mortgage rates create a housing lock-in effect that prevents homeowners from relocating. This could force employers to embrace remote work as talent becomes geographically immobile.
Read post →What Engineering Leaders Ask That Others Don't
Key characteristics that define effective leaders and mentors in software development: accountability, continuous learning, and building others.
Read post →