Software engineering

Organization, Estimations & Methodology

• Complex problems might lead to over engineering
• Discovery requires time
• Premature optimization is procrastination
• Fully implementing unknown tasks can be bad
• Task stages can be known or unknown
• There are two types of work
• One can only estimate imagined work
• Aggressive timekeeping in meetings reduces meeting time
• Glue work
• glue work is often underappreciated
• Building in public
• Learning in public

Software Design Philosophy & Simplicity

• Banana software
• Duck Features
• Don’t strive for perfect
• Cutting scope can ensures software quality
• Content history is a core feature
• Employability is mostly about your experiences
• Llms bring skill atrophy to programmers
• Schrödinger Backups

Boring software

• Software should be delightfully boring
• Principle of least surprise
• The principle of least surprise frees up mental load
• ideas to create boring software
• Last responsible moment
• Frighteningly interesting software
• Micro services should not be the default
• Keeping things simple is a skill
• Over engineering is bad for a project

Growth work

• growth work
• growth work should not interfere with base work
• growth work broadens your experiences
• growth work is compounding
• growth work should be optional and be freeing

Incremental development

• Incremental development
• Walking skeleton

Product design & UX

• Button state vs Button action
• Be resilient against feature requests
• Cutting features can improve a product
• A product should be simple and easy to use
• Features are either valuable or waste

Observability

• Logs
• Traces
• Traces are sequences of logs
• Traces can slow down a program
• logs and traces can contain PII

Structured and unstructured logging

• Structured logging
• Semi structured logs
• Tools unleashes the power of structured logging
• structured logs are hard to read as a human
• Structured logs explode in size

Architectures

• Event sourcing
• Event-driven architecture
• Collaborative filtering

Data & Data structures

• Personal Identifiable information (PII)
• Data pagination
• Offset Pagination
• Keyset pagination
• data informed, not data driven, in regards to metrics

Devops

• A successful infrastructure is one you forget about
• Self documenting systems
• Basic documentation is a must for system administration

Testing

• Test names should be straightforward
• Unit tests should be easy to read
• Mocking factories should be used when mocking test data
• API documentation can run implementation tests