Databases form the cornerstone of many applications, Web sites, and platforms. A huge amount of time, money, and research has been poured into databases over the last few decades. But our thirst for data, and the quantities that we’re trying to read and analyze, continue to grow. What can and should we do? How can we ensure reliability? How can we communicate with a growing number of other systems? And where does PostgreSQL, an open-source relational database with a growing number of features, fit into this trend? In this talk, Siimon Riggs answers all of these questions, and describes how PostgreSQL’s developers are working to keep as many of these questions in mind as they continue to improve their contribution to the world of databases.
Time: 22 minutes
As you might have heard, Volkswagen was recently caught in quite a scandal: Its diesel-powered cars would report one (legal, standard) pollution number when the car was being inspected, but quite another number (illegal and nonstandard) when it was actually being driven and on the road. This scandal, sometimes called “Dieselgate,” has had many implications for VW as a company, and for the car industry in general, But how did VW manage to fool so many people for so long? What did the car’s software and hardware do (and not do) in order to trick emissions inspectors? In this talk, Felix Domke and Daniel Lange report what they have each found, providing us with a fascinating introduction into the intersection of cars, computers, regulations, and scandals.
We often talk about “software engineering,” rather than just “programming” — but that implies that we are acting and working as engineers. The thing is, what does it mean to be an engineer, and to have that kind of discipline? And do we, in the software world, really use engineering in our work? In this talk, Mary Shaw describes what it means to be an engineer, comparing software with such disciplines as civil engineering. She then raises many questions about what we’re doing, and how we’re doing it — and how we can do it even better in the future.
Containers are a big thing in IT and devops — the idea that you can have lightweight virtual machines that can be launched, duplicated, copied, and scaled up is very attractive, and for good reason. But what happens when you have a very large number of containers? Have you simply exchanged one problem for another? Or are there methods and systems you can use to manage your containers? In this talk, Mandy Waite describes what she and others at Google are doing to make it easier to work with containers, and how you should start to look at containers if you’re to use them most effectively.
Anyone who watches video online (which is just about everyone) knows that sometimes, we need to wait for the video to buffer — that is, to download sufficiently into memory so that we can watch the video, or at least part of it. But it turns out that buffering isn’t a necessary evil; rather, it is the result of engineering and business decisions that Fredy Kuenzler believes are bad for end users, and end up costing everyone more than necessary. In this talk, we learn about the sources of buffering, the ways in which it causes problems, and some of the reasons why we pay so much for bandwidth.