Plug-Ins Considered Harmful?

For some time now, I’ve been planning something that I hope will be interesting and useful, that will be based on some sort of content-management platform.

I don’t want to say too much about it, but for present purposes imagine that it is organized a little like a newspaper, with sections each operated individually. Unlike a newspaper, here the “front page” will consist primarily of an aggregation of the “inside” sections, or perhaps of teaser versions of the full content that is actually located in the “sections”.

All content will be open for comments. Readers will be free to go straight to the “sections”; indeed if they are reading the “front” and try to comment, that will take them inside the appropriate “section”.

I want to use free tools that have a substantial user community, and do as little modification as I can get away with, in order to minimize the tech support I’ll have to do. I’d love to use Slash as the engine for this as it is very strong on the integration of a front and ‘inside’ sections, plus has lots of other community-building features, but Slash just isn’t user-friendly enough for either the posters or the commentators. Plus Slash requires mod_perl, which just won’t play nice with other things going on the machine I have available.

Currently, I’m looking at WordPress, with each section being a separate blog, and the “front page” as a sorted aggregation of the RSS feeds. [This requires a tiny bit of work, as WordPress and the most common plugins want to treat each RSS/Atom/whatever feed as a separate thing rather than a contribution to a melting pot.] WordPress is certainly friendly enough, and I like the ability that its reliance on themes give me to have a constant ‘look’ for the site but maybe vary the color scheme a bit for some sections. Perhaps WordPress is a little light on the cool tools. But then maybe I can live without the glitz. Even so, that still leaves out a few things that I think are essential.

For example, to make this particular site nice, I think I need robust and good-looking threaded comments. And WordPress, out of the box, doesn’t do that (yet). There is a threaded comments plug-in, but I’m very reluctant to commit to it because of its effects on the upgrade path. Will it work with the next version of WordPress? Who can know, when one is at the joint mercy of the WordPress developers and of the plugin writer. Or, what if a future version of WordPress has threaded comments, but in a manner incompatible with this plugin? Then I have to find some way of translating the historic comments into the new schema. (Not all this project will, I hope, be total ephemera.)

I could hunt around for a different CMS that already does what I want — I am, for example, learning about Drupal — but there are not that many with a sufficiently large installed base that I feel comfortable committing to them.

But one thing that this has made clear to me is that from the perspective of someone trying to find the right tool for a project, even (especially?) a person who wants to use open source tools, there’s a real disincentive to committing to a project that relies heavily on extensions and plug ins that are not supported by the core designers of your software package. By doing so, you not only put yourself at the mercy of there being appropriate upgrades for the main branch, but also of the plug in and, worst of all, the interactions between the two.