Conversation

The replies to this explain a lot of the interviews I’ve done over the last few years. Pro-tip: you need to look for the **earliest** discussions of the term. Recent discussions are too likely to be numpties spreading their ignorance twitter.com/clairebcarroll

This Tweet is unavailable.

This Tweet is from an account that no longer exists. Learn more

Replying to

No, not at all. I’m saying that: 1) Maybe I shouldn’t expect candidates to know quite so many of these. And 2) I often find very recent blog posts or articles that attempt to summarize a fairly old concept like “OLAP cube” are incomplete, and often wrong or at least misleading

Replying to

and

For instance, if you dig into “OLAP cubes” looking for older articles you should end up on the original 1993 paper by EF Codd “Providing OLAP to User-Analysts: An IT Mandate” or maybe Celko’s book where he provides the context.

books.google.com

Joe Celko's Analytics and OLAP in SQL

Replying to

and

I think when I was writing 'The Rise and Fall of the OLAP Cube', I'd found only Gray, Bosworth, Layman and Piramish on the Data Cube relational operator, and Harinarayan, Rajaraman, and Ullman on implementing data cubes efficiently. Celko's book looks promising, thanks!

Replying to

and

I would be interested if you find any flaws with this piece. I think if I were to rewrite it again, I would say something like "OLAP cubes used to be implemented in RDBMSes, but that's mostly fallen to the wayside because evolutionary reasons." holistics.io/blog/the-rise-

8:47 AM · Aug 19, 2021Twitter Web App

Replying to

and

And maybe I shouldn't have used the n-dimensional array example, but it's a little difficult to explain low-level data structures to data analysts.

Replying to

and

Ugh, another thing I could have made clearer: 'data cubes started out as a visualisation thing, but then because most modern cubes are tied to a custom data structure implementation, 'olap cube' today refers to a type of product that isn't db related.'