Faster Than Dijkstra?
3 months ago
- #algorithms
- #routing
- #networking
- A new algorithm claims to improve on Dijkstra's method for finding shortest paths in networks by avoiding sorting operations.
- The new approach has been peer-reviewed and published, but its practical impact on real routing systems is questioned.
- Scaling limits in routing systems may not benefit from theoretical improvements if the network size (n) isn't large enough.
- Historical examples show that scalability isn't always the primary factor in successful commercial products.
- Routing performance depends on multiple factors beyond SPF calculation, such as failure detection and convergence times.
- Dijkstra's algorithm remains popular due to its simplicity and understandability, making it preferable for implementation.
- The article also touches on unrelated topics like datacenters in space and AI-generated content on Wikipedia.