Ace Roqs’ Blog

Last Sunday, I went feature complete on TurboHex 1.5.  1.5 has a checkered history, and it’s lengthy development mirrored a lot of my external activities throughout it’s development.

The code that eventually became 1.0 began in 1999, when I started experimenting with high performance disk and file editors.  Full development took place in the Spring of 2002, and was released in the Summer of 2003, after spending many months writing, and then throwing away, custom e-commerce software.

Most of the features in 1.5 were added to the product in the Fall of 2003, but the code base fell into decay after Dead Man’s Hand took over my life, and then Microsoft after that.  2004 saw almost no development, and the latter half of 2005 was the same.

The first half of 2005 involved an number of lateral changes, as I moved the code to Visual C++ 8, added safe string support, added support for x64, added support for GCC, added support for doc comments, and other changes that really didn’t do much except churn a lot of code.  The result at the beginning of 2006 was a lump of code that was in serious need of refactoring, as multiple features were in half-finished and buggy states, and the massive amount of code churn caused various instabilities.

I found myself in a similar state as the end of 1.0, where I had to make a decision to cut features in the interest of actually releasing something.  1.5 has a good feature set, but I always wish I could add more, and there are a few pet features that I am really disappointed didn’t make the cut.

Over the last week, I’ve been fixing bugs and adding the polish that is required of a final product.  Sometimes this can be boring work, but this has been fun as this is the first time in a while that the product has really good quality again, and things that have been bugging me for some time are now actually fixed.

Even the small features, such as the color of the editor stripes have been tweaked.  The 1.0 colors were developed on my laptop, and as such with LCDs, they looked great there, and horrible everywhere else.  Now the colors just look great.  Sometimes it’s the simple things.

The user interface is still streamlined, and that is something I really take pride in, but some people think you that the more complicated and cluttered the UI is, the more powerful the app is.  I disagree - and that’s why it looks like it does.

I’ll be finishing the last set of bugs and help file fixes shortly and then sending it to my beta testers.  Users will now have a configurable MRU (most recently used file list), process editing, partition table viewing, more powerful disk seeking features, and dozens of other bug fixes.  It has been too long for this feature list, and I have learned a lot about keeping things in a shippable state.  Hopefully future releases will be far more rapid, and that will enable a broader feature set.

It is still the most compact yet powerful hex editor around.