On Mike Cannon-Brookes’ Open Source
On Mike Cannon-Brookes’ Open Source J2ee Landscape. This is very useful to me. A simple list of the many Open Source J2EE projects he knows about, and his favourite components from them. It’s a lot easier to get a basic feel for what each project is about from Mike’s two-line summary, rather than by trying to find, read, normalise and compare each project’s description of itself. While Mike admits it’s a subjective view, I still feel it’s a good idea to be familiar with all of these components’ existence, purpose, and development status, lest you allow your blinkered dev leader to walk you into coding something identical in purpose, and inferior in quality.
So this list is useful now, but how do I know what’s missing from it today, or in 6 month’s time?
The human summary needs a human author, but the list itself of top components could be generated. If every successful or partially successful j2ee dev project presented a structured list of the components it built with, and perhaps also simply documented the experience it had with them, a list like Mike’s could be generated, to include the components that repeatedly proved themselves. And there could be a set of user-opinions on each in the list. Indeed, a list of the components with highest profile which also ran the project team into the greatest difficulty could also be generated and would be at least as useful. “Ah, it seems everyone who uses this object-relational mapping software X has real config problems with database Y / says performance is unnacceptable / says the documentation is bobbins”.
Where does the list get hosted (Mike says he can’t find the info anywhere else today, which I guess is why he compiled the brain dump), and who decides the structure?
The manual alternative of this is to do the same research by finding and reading a lot of project post-mortems, whilst all the time wishing them into existence.