Where the hell’s that software? Where’s that CMS Kevin keeps going on about and apparently has been working on as long as… Forever, it seems?
Good question indeed, and one that needs answering. What happened so far was that once I realised that one of the really interesting things on the Internet was open information systems, rather than clever interfaces in Flash or design or anything like that – anyone can design some clever buttons or make a font too small to read, but designing good information flow is really very tricky. And that hooked me. I found out about PHP, and MySQL, and Apache, and I went from there. Slowly.
Then, I started a new job in Parallel (where I still work) and saw that what these systems are is Content Management Systems. I managed to build the bonesd of one before I went travelling in 2003 but it didn’t work (in fact, it was hideously bad). However, while I was away I had a chance to study some Java and I realised how these things should be done – not with PHP, at any rate, I decided. For those of yolu reading this and who know what PHP is: yes, PHP is good for certain things – anything you would do with ASP, for example, and probably a lot better – but it isn’t an object-orientated programming language, and it’s messy in a way that I can’t stand (there are functions for just about anything you can possibly want to do, which is great, but no consistency and a lot of doubling up because two of those beardy little open-source guys hiding away in their pods each thought of some idea at the same time but didn’t tell anyone). Of course, I didn’t have the Java skills to get stuck straight in and I built the first version, which is curveLight, in PHP and it wasn’t too bad for a first shot; the idea being that I should build a proof of concept and get the architecture down first. The thing with a good OOP architecture is that it should be to some degree language agnostic. If the design is good then it will work anywhere.
The design of the architecture has been thrashed and trashed a few times since then, and apart from time constraints I’m ready to rock and roll with the design as I think it should be done, and when I’m satisfied that it works in PHP then I’ll port it to Java. The current thinking is towards lightweight, a blogging-friensly tool rather than something heavy, but if it’s a good design and extensible then it will scale well anyway. This will be… I won’t say. It has a name.
After that will be the big Java version, with more heavy-weight capabilities, to tackle many users, concurrent sessions, geography, media management, instant messaging, all the usual good stuff. It’ll be hell. I’ll move to a small hut in a remote area and work in isolation for years, I suspect that’s the only way. Then I’ll come back and discover that everyone’s using a new technology called ‘Sphincter’ which is deployed on NYAK networks.
Incidentally, I intend to Open-Source all of it and spread goodness in the world, Everyone Should Have One.©
In case it isn’t clear already, this site is currently being run on WordPress – which is stable, simple, easy to use, and fiendlishly clever at the same time. It’s PHP, MySQL and Apache based blogging software which free and Open Source, both of which are very good things. I was resisting it for quite a while but now I’ve taken the plunge, and I’m wondering why I should even bother developing my own…
[edited 16th September 05]