I år er første gang JavaZone har blitt utsolgt for billetter! Dette synes vi er en veldig spennende utvikling, spesielt siden vi har ansatte som har bidratt med både innhold og organisering av arrangementet i flere år. På foredragsfronten stilte vi i år med en workshop, tre foredrag og to lyntaler. I god Computas-ånd ønsker vi å dele vår kunnskap med de rundt oss! Uansett om du ikke fikk anledning til å delta på årets JavaZone, ikke fikk sett våre bidrag, eller vil se de om igjen, kan du finne de samlet i denne posten.
Rustam Mehmandarov og Khiem-Kim Ho Xuan
How We Built an Indoor Map For the JavaZone App. You Can Do That Too!
You just zoom into the map until you start to see an indoor floor plan of the building. Then you use the level switcher to move between available floors in the building. Isn’t that cool? Well, obviously it is. But it is also really useful. Have a look at what can be created when a map geek meets an Android geek. Oh, wait, you probably have seen the results already in the app. Well, come and see how we did it, and get a tour behind the scenes.
Filip Van Laenen
Dial M for Mutation!
Mutation testing is the true test for whether or not your source code does the right thing, and whether it’s properly tested by your unit tests. But while we’re on a killing spree against the mutants of our source code, let’s reflect a bit on what these mutants really represent, and why we want to include the unit tests that kill them in our test suite.
What’s the real point of boundary testing? Do we really need every statement in the source code? Have we tested all calculations? How do we know that we got all the access modifiers on our methods right? These are some of the questions that we would like to reflect on for a moment.
We’ll start this talk with a short introduction to mutation testing, but only to set the stage. The core of this talk consists of a walk-through of Java code and unit tests, and a reflection on TDD and the quality of our unit tests. At the end of this talk you should be more suspicious about both your source code and your unit tests.
Disclosing patterns and phenomena in crime with semantic Java!
«Hendelsesregistrering» (Event Registration) is a prototype developed for the police in Norway, focusing on assisting crime investigations and analysis by being able to detect phenomena and patterns in crime. The aim is to give the police the information they need at the right time to make better decisions, support crime investigations, see connections between events, and identify phenomena and patterns in crime.
The prototype is based on semantic web technologies, and adopts the Apache Jena framework for semantic web programming with Java. This presentation will go through the functional aspect of the solution, with a closer look at how the Apache Jena framework for Java supports the entire back-end stack.
Learn to be Lucky!
TDD is all good but testing is more than checking that the code does what the requirements says it should do. OK, it might happen that we forget a requirement or two. However, the majority of bugs arrives in situations that are not mentioned in the specfication. How do we discover these situations? Luck? Going back to waterfall and spending months on getting the requirements perfect before we start coding? Performing rigerous manual testing after the development is done?
Exploratory testing is a powerful approach to testing that has been used by the testing community for more than 25 years. It is especially useful when working with complex solutions, incomplete requirements and specifications, and/or time-pressure. Does the situation sounds familiar?
In this talk we’ll look at how to apply the ideas behind exploratory testing and use them to find potential error situations and write more robust code. You might not learn to win the lottery but you’ll learn how to find more of the nasty bugs before you commit your code to the repository.
Ikke gjør som kunden din sier!
I det siste har brukersentrert design og UX blitt det hotteste hotte som er på alles lepper. Men hvis vi bare gjør som kunden sier, glemmer vi at en viktig del av rollen vår som eksperter, er å stille spørsmål ved ønskene som strømmer på.
I denne lyntalen ser vi på hvorfor det lønner seg å ikke gjøre som kunden sier. Vi ser på noen skrekk-eksempler hvor kunden fikk bestemme litt for mye, og tilfeller av det motsatte – hvor løsningen ble mye bedre av å ignorere brukernes ønsker.