I've decided that I really should start "diarising". I'm sure it will be a good form of self-motivation, if nothing else... (ie even if noone actually reads what I say here)
Been working recently on a "product comparison engine", for want of a better term... I had been thinking of doing it for a while, based on a little business idea thing I've got, and I had written out some db schemas etc, but only when Cam Barrett (I don't think he's on Advogato?) started asking for help on the cms-list did I start writing code.
Unfortunately, Cam wanted the code written in PHP, so I had to throw away my utopian ideas of a real J2EE-compliant MVC-based architecture, and hack up something ugly in a language I was teaching myself on the fly.
Some thoughts on PHP
Personally, and selfishly, I don't see the point in a language like PHP.
What I mean is this: all the docs out on the Net that proselytize PHP keep on talking about "how easy it is to learn, even for non-programmers". Sure, so is BASIC. And it teaches you lots of bad habits, unless you are very careful about what you do. And, like BASIC, some things just can't be done cleanly (like separating display logic from the display itself). They even talk about how bad Perl is for new programmers, and how nice PHP is compared to Perl because it doesn't look like line noise, but then they boast that PHP is compatible with Perl regular expressions!? Go figure.
Anyway, I got most of it finished in PHP and was going to hand it over, when I had an interesting conversation last week with some people that might be interested to help me maintain the system myself. As I've just started up a content-management consulting company, it would make tons of sense for me to do it myself, and frankly I think it would be better for the CMS community if I had direct access to my code so I could roll out new features without having to go via Cam (and without having to use a crappy scripting language)
The upshot of all this is that I'm currently thinking of rewriting it all in Java servlets/JSP/Velocity (true MVC, or as close as you can get right now), hosting it myself somewhere, and promoting it like crazy. This would also give me the opportunity to create other product-comparison type sites using the same code base, and possibly release the code as open-source.
So now that I've committed (at least to a Web site?!) let's see how far I can go.