All in the <head> – Ponderings and code by Drew McLellan –

The State of Textpattern

This site is published using Textpattern, and has been for more than four years. In fact, save for a few of Dean Allen’s own sites, this is pretty much the longest standing Textpattern installation going. I jest not. A major reason for starting this site at all was to play around with this new toy Dean had given me to try out. Back then I was running PHP as a CGI, which exposed a number of compatibility issues and helped iron out a few problems long before Textpattern was released to the public.

It’s not been a perfectly smooth journey. From the get-go this was always something Dean was working on his is limited free time, and releases would come in fits and starts. Even once Textpattern was released to the public, updates would sometimes be months apart. All for good reason, and security fixes were never neglected, but you know, months.

Eventually, of course, the inevitable happened and Dean had to hold his hands up to really not having the time to work on Textpattern any more. By this time the code was already GPL’d, and so it was handed over to the care of the already active developer community. Even still, releases were months apart.

Never about features alone

Textpattern was always intended as a tool to enable the author to ‘just write’. (Hey, that would make a great slogan MyPublishingTool: Just Write.) It wasn’t about having the most features or templates or plugins, but simply about being a good, elegant tool for personal publishing, primarily for weblogs. It covered the basics of weblog publishing, and in a politely opinionated way. (For example, Textpattern never had a calendar feature for accessing archives, the reasoning being that it makes no sense to access posts that way. Very rarely do you care what someone wrote purely by date alone.)

As the months in between Textpattern maintenance releases went by, not only were other publishing tools making significant steps forward, but so was the entire concept of personal publishing. Weblogs were evolving, and still are evolving. With the exception of the plugins system, the Textpattern that I use today is fundamentally the same beast I was running back in the March of 2003.

As I said, Textpattern was never about being rich in features. However, there’s a certain baseline of features required for a tool to be useful, and for personal publishing software that baseline is ever rising. Today, I count features such as robust comment spam detection and tagging as essential features. OpenID will be on that list within 18 months. Whilst Textpattern’s plugin architecture is good, there’s a limit to what can be done without beginning to mess with the underlying database structure. You can only allow a single plugin to change your schema, because after that no other plugin knows what it’s going to find.

So what’s the state of Textpattern?

Textpattern, now in the hands of a small team of community members is not dead. It does, however, appear to be in something of a coma. The most recent release – a mere double-dot release – was 7 months ago. There’s an ‘experimental’ branch in SVN, but with no clear aim or goals. There is no roadmap, because the team aren’t in a position to commit to dates. Why they can’t commit to features without dates is unclear.

The team are, however, happy to use the official project blog to pimp their commercial plugins. Well, ok, you give a little you get a little – we all have bills to pay. Seems like they must have a lot of bills to pay, as the community are now being asked to sponsor further development. Of what, we’re not told. There’s no roadmap after all.

So apart from being a bit disillusioned with the state of things, where does that leave me? Well, I have a refreshed design that I want to roll out here. I’d like to be able to tag my posts in a meaningful way. I need some way of fighting comment spam, and, alongside that, I want to use OpenID for comments. There’s a few dozen other things I’d like to do, but those are the basics.

I’ve toyed with the idea of building my own system, but don’t have the time. I don’t really have much time to contribute to the Textpattern project, although I could if I knew what we were working towards. With the current state of that project, there’s no guarantee that any invested hours will be worthwhile.

So I’m thinking of switching to WordPress. As much as I love Textpattern, the long term prognosis isn’t good.