Our latest podcast is here. The topic this time is JavaZone – the biggest community-driven conference for Java developers. The conference took place in Oslo last week and four of our colleagues gave talks, and two were leading a workshop. In addition to that, we were one of the partners and had our booth at the expo area.
In this podcast Filip Van Laenen, Cecilie Haugstvedt, Mads Opheim, Rustam Mehmandarov and our guest from Google, Ray Tsang, will give a brief overview of the conference, our contributions to it, as well as other highlights and general impressions.
This year at JavaZone we presented three lectures and one workshop. In good Computas spirit, we want to share our knowledge with those around us! Whether you did not get to participate in this year’s JavaZone, did not see our contributions, or just want to see them again, you can find them in this blogpost.
Teodor Ande Elstad og Simen Haugerud Granlund
Lag din egen sjakkcomputer og revolusjoner verden!
This was a norwegian workshop, therefor no video. The description is in norwegian too.
Kunstig intelligens holder på å revolusjonere verden. Alt fra selvstyrende biler til automatisk kassebetjening i matbutikken er med å gi store omveltninger i mange bransjer, og dataprogrammer utkonkurrerer mennesker i stadig flere oppgaver. Ord som som nevrale nett, maskinlæring og intelligente agenter kastes rundt, og det kan være vanskelig å vite hvor man skal starte for å lære om kunstig intelligens.
I denne workshopen tar vi for oss det grunnleggende, og du vil få en introduksjon til en del fundamentale temaer innen klassisk kunstig intelligens. Det du lærer vil du bruke til å lage din helt egen sjakkcomputer og konkurrere mot de andre kursdeltakerne.
Filip Van Laenen
Clouds with Trenches and Sharp, Bleeding Edges
So you’re going to run your project in a public cloud. That should be easy, because everything is taken care of, “as a Service”, right? Well, think again, because it turns out clouds have trenches too. And sharp, bleeding edges.
Sure enough, you don’t have to set up any hardware. But you still have to set up virtual machines, nodes, containers, or whatever it is you’ll be deploying your software to. This will have to be configured in a development environment. And a test environment. A staging environment. A production environment. And a couple of environments more, because it’s easy and cheap to create more environments, right? You better think that through from the start of your project.
Then there’s security. Not just the firewalls that have to be configured, but authentication and authorization. Your data isn’t just sitting around on a local server in an internal zone protected by your corporation’s firewall. It’s in the cloud! Uploading data to the development environment is suddenly the same thing as running a production system.
Performance becomes a new challenge in the cloud. You probably won’t be able to outscale Google, Microsoft or Amazon, but they can pretty easy outscale you with a large bill.
And finally, the joy of beta versions, or if you’re a privileged partner, alpha versions! Are you sure you want to touch them? Or can’t you afford to not touch them? You’ll be bleeding anyway.
Developing software in and for the cloud is a lot of fun, but don’t expect to always be on cloud nine!
Anne Landro og Mads Opheim
DDD + legacy-monolitt = sant?
This talk was in norwegian, therefor the description is in norwegian too.
Har DDD noe for seg på et legacy-prosjekt? Funker teknikkene også på 15 år gammel java-kode? Kan vi bruke DDD på en avgrensa del av en monolitt, eller er det bare brukbart for microservice-prosjekter? Er det umulige mulig?
Vi stilte oss disse spørsmålene, og bestemte oss for å finne det ut ved å prøve i vårt komplekse prosjekt. Hva ville skje hvis vi ga DDD en reell sjanse? Hva funka og hva funka ikke, hvor møtte vi veggen, og ga veggen etter?
Vi deler våre erfaringer og problematiserer bruk av DDD så vel som manglende bruk av DDD, og hvilke feil du ikke trenger å gjenta etter oss.
Build the right system / It’s not enough to build the system right
There are a lot of technologies out there that will help you determine if you’re building the system right. That’s good but really no help at all if it turns out you’re building the wrong system. If you really want to speed up development and deliver value to your customers as fast as possible you need something more than another language, framework, platform or library.
How can you make sure you’re building the right system before you start coding? How can you get fast feedback on what you’re building while developing? How can you check that you’re delivering actual value to your customers before you send your code to production? How can you get more time to play and eat cake in the office?
Subjects include shift-left, personas, BDD, mob programming and exploratory testing.