On Leaving Habari

I was asked by Michael Harris why it was I planned on leaving Habari, so here goes:

1. Autop is the wrong level of abstraction. In the grand scheme of things you’re either writing HTML or you aren’t. “Aren’t” can include a lot of other things that have defined transformations into HTML (such as Markdown) but you’re still selecting a single format to compose in. Autop falls in the uncanny valley of of composition. It functions just fine for most posts and then makes the world blow up in other scenarios. I have a number of entries which are single lines of HTML to work around the autop functionality–which I can’t disable because other entries of mine rely upon it. I didn’t know that autop was going to end up being such a mess, but at this point I fully believe that having an RTE where you can toggle between editing source and editing content is the way to go.
2. I make a living writing code, but that doesn’t mean that I want “the tool on which my website is built” to be one of my side projects. I truly believe that Habari’s code quality really sets it apart from WordPress but that has made zero difference to me: I haven’t made any changes to Habari since setting it up three years ago. I fully expect the same thing to occur once I’ve set up WordPress. There are just too many other things I would rather spend my time on.
3. Habari is a wonderful blogging engine. WordPress aspires to be just a little bit more than that, in spite of it not succeeding particularly well at either. I’m planning to turn it into a CMS for managing client work in addition to it being my blogging engine and I’ll deal with its shortcomings as I run into them.
4. Three years ago the admin page for Habari was amazing and beyond anything I’d seen from other blogging engines. There was a tremendous excitement in the community and development was moving at an incredibly rapid pace. The code was cutting edge for PHP and doing everything right. Cut to three years later there have been three official releases: 0.7, 0.7.1, and 0.8 (I didn’t research the timeline, and I may have updated along the way, but I’m currently running a 0.7-alpha trunk build). The features which once set Habari apart are now more mainstream.

I don’t yet know what features I’m going to miss, but Habari has treated me really well for three years and I leave with no hard feelings. I hope that the momentum from releasing 0.8 will get Habari all the way to a high-quality 1.0 release. And besides, I know that Habari will always maintain a script to switch to it from the current release version of WordPress. :)