Coding After Work Podcast

RavenDb, passion, and books with Dejan Milicic

February 15, 2024·42 min·3 clips
Dejan Milicic explains why coding privately remained enjoyable while workplace coding became difficult.
1. Coding After Work Podcast examines RavenDB, passion, books, and software development with Dejan Milicic. 2. Dejan Milicic works on RavenDB product demonstrations, educational sessions, and workshops while still writing code without release deadlines. 3. The episode asks how pragmatic database design and personal learning can reduce development friction. 4. Dejan traces enterprise software in the 1980s to internet applications whose potential audience became the whole planet after the mid-1990s. 5. Dejan connects Black Friday website crashes with the need for database approaches beyond traditional relational systems. 6. Google, Yahoo, and Pinterest developed specialized solutions for problems such as indexing the entire internet. 7. Dejan describes NoSQL databases as pragmatic responses that relax relational-database conventions and reduce object-relational mapping problems. 8. Dejan discusses temporal modeling for purchase orders that preserve customer-specific prices and discounts at a particular time. 9. Ted Neward's polyglot programming idea connects language and database choices to different services, including Python, Node.js, relational reporting, and Elasticsearch search. 10. Dejan presents RavenDB as community-driven, with its roadmap shaped by listening to actual users rather than only following trends such as vector databases. 11. Dejan says RavenDB 6 combines evolutionary improvements with a revolutionary new indexing engine. 12. RavenDB 6 sharding distributes database pieces across servers while preserving a seamless application experience. 13. Dejan places the practical sharding use case north of a 10-terabyte database, which he estimates applies to about 1 percent of teams. 14. CoreX was built from scratch after the RavenDB team studied and optimized Lucene.net over several years. 15. CoreX processes document batches and stores indexing structures on disk instead of allocating all of them in memory. 16. Dejan says this design addresses slow cold queries and large memory allocations associated with Lucene.net. 17. Jimmy and Jessica conduct an energetic conversational interview that moves between RavenDB 6, coding habits, books, and technology history. 18. The hosts use personal examples, including a five-minute map-radius feature and previous RavenDB workshops, to test Dejan's technical claims. 19. Developers exploring RavenDB, NoSQL, and technical writing. 20. Listeners uninterested in databases, programming, or developer books.
Listen to the show on