Software is too expensive to build cheaply…

In the first installment, I outlined a simple web application. In this post, I will be turning that application into two parts – an OSGi bundle, with a client web app.
(more…)

Tags: java osgi tutorial

I’m going to be spending some time over the next couple of weeks learning about OSGi – there is an application at work that we want to try and make more modular. In particular, we’d like to be able to share the back tier with more front end clients. The more conventional modularisation techniques, such as EJBs, have been tried and didn’t work fairly well. Simply creating more deployments is prohibitive, due to hardward And before someone asks “why not just stick a web service on it and share that way” – some of the front end clients will be those web services. To cut a long story short, one of the options we want to check out is OSGi.

The only problem is that there isn’t much in the way written up on the web about writing OSGi components – at least not without invoking magic tools (e.g. Maven plugins) that don’t work on any version of any tool developed six months later. Which is odd, because OSGi, as a spec, doesn’t look that hard to write too. So, I thought I’d write up a series describing the initial investigative spikes, starting as close to the metal as possible and working up.
(more…)

Tags: osgi java glassfish

I saw an an article on Seth Godin’s blog about positive thinking. Godin makes the valid point that people who think positively tend to succeed more, in part because their confidence means they don’t second-guess themselves. You can easily waste a lot of energy debating what to do.

A point that Seth didn’t cover is that positive thinking means it’s more likely you will attempt something new: if you feel you can accomplish a task, you’re more likely to try. Sometimes, when you try something new, you will succeed. Other times you will fail. When you fail, there’s a chance you will learn – and learning makes it easier to succeed next time.

One of the key take aways of agile development for me is “experiment, and fail early”. Fail early, fail often, fail cheaply. And learn every time.