I think that a common occurrence on HN, which I'm definitely a perpetrator of, is upvoting things you wish you had a better grasp on. Theory-heavy & math posts seem especially prone to this.

Perhaps people find it interesting but are too shy to leave the first comment. Like being the one to ask a question during a lecture, sometimes it's really hard to put yourself out there.

I'm curious too. As a CS grad student with relatively little database knowledge, would this be a good text to start learning?

I haven't fully read it (yet), but based on a quick scan of a few chapters, it looks like a dense but well-written introduction.

Some additional resources I'd recommend if you're interested:

Designing Data-Intensive Applications is a fantastic place to start if you're interested in the intersection of databases & distributed systems:

https://dataintensive.net/

The Architecture of Open-Source Applications book has a fewer chapters on databases:

https://aosabook.org/en/bdb.html

https://aosabook.org/en/hdfs.html

https://aosabook.org/en/nosql.html

There's some fantastic documentation on Postgres and its internals:

http://www.interdb.jp/pg/index.html

https://momjian.us/main/presentations/internals.html

https://www.postgresql.org/docs/current/internals.html

Thanks so much! I'll definitely take a look into these.

On the preface of the book (omitted from the ebook), it details several learning paths:

The Classic: a basic database theory course covering the classical material would center around Parts B and C. Chapter 10 and parts of Chapter 9 of Part C are somewhat more advanced and could be skipped. If time allows, some of Chapter 12 and a selection of Part F might be covered.

Feast of Query Languages: a course on the theory of query languages would start with a quick review of the basic material on classical languages (Part B), and continue with Parts D and E. If time allows, some material on languages for complex objects and object-oriented databases (Part F) could be covered.

Gourmet Sampling of Database Theory: a course for people with theoretical appetites that emphasizes the specificity of database theory. Logicians wishing to explore the connection between finite-model theory and databases will be interested in Parts C and E. Those interested in descriptive complexity will find Part E closest to their hearts. Researchers in logic programming will prefer Part D, particularly Chapters 12, 13, and 15. People with a background in theoretical artificial intelligence will find Parts D and F of particular interest. Rule-based systems are related to Chapter 14 (see also parts of Chapter 22). Programming language people will be interested in much of the material on query languages, including Chapters 20 and 21 in Part F.

Fast-Food Database Theory: a course for applied database students that is meant to provide informal exposure to some of the basic results of database theory. This would include assorted topics, with an informal presentation emphasizing the examples, results, and intuition provided in the text, rather than the proofs. A possible syllabus would include Part B; parts of Chapters 8, 9, and 11 in Part C; Chapter 12 and parts of Chapter 15 in Part D; and selected chapters of Part F.

Numerous exercises have been included, and they are essentially of three categories. Routine exercises (unmarked) require easy manipulation of the basic concepts and results. Harder exercises are marked with a (*). Another category of exercises is meant to complement the material in the text and often contains results from related research articles. These exercises are usually on the hard side and some may constitute term projects. These exercises are marked with a ( ).

With databases or anything in general, I believe in doing first and then learning the principles.

Play around with a small RDBMs like sqlite. Do a project around it. Look at the source maybe and then learn about the principles. Foundations of databases are based on discrete mathematics ( where relational comes from in RDBMs ) so a background in discrete math helps.

For enterprise level stuff ( clustering, replication, distribution, etc ), it's more practical to learn on the job where you have access to the hardware, servers, etc.

This is a text for understanding the fundamentals of databases but might be too deep if you just want to use an existing database.

This is really interesting. Now it's ranked #20 with more than 500 points, and it now has 36 comments. But none is actually about the link. People are giving advice on database books they read, and arguing about which one is the best, but it looks like nobody read this particular one. Yet it's massively upvoted.

That's not surprising. People often upvote an interesting conversation regardless of the actual article link.

In fact I added this thread to my favorites because the book references look useful.

I think a lot of people use the upvote to bookmark a link. The favorite feature was rolled out much later and I've always used votes for that purpose.

I have seen it previously and it reads really well :-)