I really like the idea of an open-source genealogy program. The only real open-source genealogy program that runs on the desktop seems to be GRAMPS. GRAMPS stands for Genealogical Research and Analysis Management Program System. Yeah, I'll stick with GRAMPS. Originally developed for Linux, it now has versions that run on Windows and on the Mac. I've tried it in the past but never really gotten it to work. When I saw a new version released today I installed it and tried again.
Unfortunately, I ran into error messages right from the beginning. I suspect this has to do with running the program on a Mac, a platform they claim to support, but which is not supported very well. After I got past the initial error messages (which prompted me to submit a bug report, but the bug reporter feature led to an error as well) I tried to import a GEDCOM, but found that the Import function didn't do anything.
I suspect GRAMPS run significantly better on Linux, but alas most people don't use Linux. As with most open-source projects, the development of the project is directed by those developers who choose to do the actual work. In the beginning, all the developers of GRAMPS were on Linux, and the goal was to create a genealogy program for that platform. It is common sense that under such a situation Linux would be the focus. Over the years GRAMPS has added developers who have chosen to make GRAMPS work on other platforms.
In a recent blog post by one of the developers in response to a post in their own forum about a two-year old blog posting on GRAMPS for Windows, the general ideals of this (and many other) open-source project is revealed, that the developers decide what is best, and if you don't agree then become a developer for the project and then you can't have an opinion. In this case, the original two-year-old blog posting was recommending a very specific technical route to take, which I would agree if you're not going to contribute to the development efforts, is probably a silly thing to do. Developers don't listen to outside suggestions - they listen to inside suggestions, so even if a developer working inside the project made the same exact technical suggestion, it would be better received than from an outsider. Why a developer decided to bring this up two years later, when the technical issues being discussed are very different than they were in 2008, however, is beyond me.
I would agree with the original sentiment of the two-year-old posting, however, that if you are going to port your application to platforms like Windows and Mac, you should be prepared to use the interface tools available natively on that platform. The look of GRAMPS on the Mac, while familiar to a Linux user, is totally foreign to a Mac user. If I was going to make a suggestion to the GRAMPS developers, it would probably be to look at a cross-platform GUI toolkit like wxPython which lets one use native GUI widgets for each platform, so on Windows the buttons are Windows buttons and on the Mac they're Mac buttons. If they really wanted to support the Mac natively they could look at PyObjC, but that wouldn't help them for Linux or Windows, so wxPython is probably a better choice. Of course, I'm not a developer, and I am not offering to become one, so my opinion doesn't really count here.
The hopefully good news here is that in that same developer blog posting, he says the forthcoming 3.3 version has the best support for multiple platforms GRAMPS has ever had. It's planned to release in March, so I guess we'll see then how GRAMPS has improved. When GRAMPS 3.3 is released, I expect to try it again, and post the results here.
Anyone reading this using GRAMPS? I'd love to hear your thoughts in the comments.
[June 27, 2011 - I've posted a follow-up to this post on a newer version of GRAMPS, 3.3.0, which is improved in terms of stability, but is not yet ready to use on a regular basis on the Mac in my opinion.]