This week has been about Rust, a knowledge database, OLAP cubes, security issues with CSS and techniques for fuzzy matching.
A graph database focused on knowledge graph representation. The new version, 1.1, introduces a new storage system with revision control à la git. In their words:
We adopt a delta encoding approach to updates (‘like git, but for data’) which provides the whole suite of revision control features: branch, merge, squash, rollback, blame, and time-travel facilitating CI/CD approaches on data.
In almost fifteen years with an on-and-off relationship with RDF, TerminusDB is the first database that I try that doesn't force you to manage all the complexity typically associated with RDF. I believe they have found a plausible abstraction for people that want to work with graphs without investing in the guts and bowels of RDF.
A comprehensive list of trade-offs on the topic of maintaining a local copy of remote data you don't own (or rather, you may own but you don't control).
CSS Exfil is a method to let an attacker steal targeted data using CSS alone.
Mike Gualtieri explains in detail how the attack works and provides extensions for Firefox and Chrome to mitigate it.
This article goes through a couple of methods and their strengths and weaknesses to match names.
This is the first part of a series of articles explaining what trade-offs make Rust be slow at compile time. The short and oversimplified answer is: fast run time and strong safe guarantees.
This article may interest you in two ways. if you are not familiar with the concepts of OLTP and OLAP it will serve as a lightweight introduction; otherwise it will serve to understand why OLAP cubes are less relevant than they were a decade ago.