Staying On Track - Sun, 11 Sep 2011

I’m continuing work on the documentation. While doing that, I’ve noticed a few small bugs in the program and I’ve been correcting them as I go.

But I’m also being careful not to get too carried away. First I wondered if there was too much overlap between the Tutorial information and the Reference section. I knew I had a lot of detail in the Tutorial about the Everything Report, and I wondered if I was repeating myself and rewriting what was already written. But as I read back through the tutorial, I satisfied myself that it was more of an instruction guide to follow for new users, whereas the Reference section was for specific information on what everything does.

Then while documenting the Notes Details and Object Details sections, I looked for an example to include among my test GEDCOMs, since the two sample files included with Behold did not produce those sections. When I found one, then I also noticed a slight problem in it. The source citations were not sorted perfectly withn the source they belonged to. The reason was they were not using the Smart Sort that I was using everywhere else.

I thought that would be easy to fix, but it isn’t. I could do it one of three ways, and the best way was to make some changes to Behold’s data structure. They weren’t too difficult, but they’d take a few days. (A few days to this programmer usually expands into an unintended few weeks.) I thought about it overnight, and concluded that its better to not tinker with Behold’s data structure now, no matter how small a change, just prior to Releasing version 1.0. Behold works well and its current data structure has been well tested since the last set of changes to it during the beta.

Why introduce possible problems now? I will be making some data structure changes to Behold to save to its own file format which is the next step after 1.0, so it’s a natural fit to make this change then. The key is staying on track now, which I’m trying to do.