Reinventing the wheel... as a service
summary
Argues that software reinvents the same solutions and shows concrete failures: XML→JSON→YAML leads to surprising parsing bugs like YAML 'no'→false; machine learning finds spurious correlations that need to be filtered with symbolic rules; architecture swings from mainframe to cloud to serverless as mainframe and back, and JVM threading cycles back to virtual threads after actors/reactive experiments. Recommends checking prior solutions, spotting oversimplification, and combining old and new tools rather than relabeling the same design.
AI-written summary · about
chapters (11)
- 0:00Start
- 1:20Hierarchical DB vs. Relational
- 6:51Fallacies of distributed computing The network is reliable
- 9:24Why not Machine Learning + Symbolic Reasoning?
- 10:46Pendulum of innovation
- 13:05Spiral of innovation
- 16:57DON'T TRY TO SET UP A SERVER
- 17:13Remote calls/ inter process communication
- 17:47Modularization
- 18:55Software Engineering / Architecture "innovation" patterns
- 19:54and what you can do about it
related talks
36mReinventing the wheel ... as a service
33mFrom Enterprise Architecture to Software Design - Critical Thinking
50mTactics and strategies on software development: how to reach ...
57mStop writing software like it's 1999!
57mBuilding Systems that are #neverdone
54mSoftware architecture evolution, what to expect? Panel Discussion
43mHow to Innovate in a Legacy
39mSoftware is a meal best served with different architectural spices