Dogfooding is good for your software. Nothing seems to sniff out bugs more than just regular usage. Most open source developers don't have a testing team, and BeOS and Haiku developers don't have a huge community they can depend on for testing. Don't get me wrong, gentle reader, I'm not ripping on the BeOS community. By no means. The community is just not as big as the Linux hordes or the marketplace for Windows and MacOS. This means that while a particular program might be quite good and appealing, there just aren't that many people downloading BeOS software. This would be why I've been flinging code with the IDE of my own making.
Writing code with Paladin has been quite nice, actually. Once in a while I'll run into something that makes me say, "Wait, it shouldn't do that," or "What in the world?! Why doesn't that work?!" Some of the features that I've put into place are nice to have, like the Run Logged or the ability to set run arguments. These features -- or ones like them -- might be commonplace in Windows land, they certainly aren't here.
What have I been working on, you ask? Minesweeper, of all things. This is another little time waster that I've enjoyed on other platforms and, like BeIDE, there isn't a version that can be legally redistributed. It's playable and is skinnable, but it's missing some stuff -- high scores and a few other smaller features. It'll be nice to have around once I have it ready to release.
As a result of all this hacking, Paladin is, once again, a better IDE for wear. There is a major regression in RC5 which causes empty files to appear in the editor when you double click on a file in the project window. There are some other bugs that have gotten a good squishing, but there are a few more items on my to-do list before it's quite ready for release. Despite the seemingly endless number of candidates, there will probably be a point where I declare a 1.0 despite outstanding bugs, but in a case like that, none of them will be showstoppers like the one I just described above.