The onslaught of comment spam on my Movable Type blog finally made me scream uncle: I switched to WordPress this weekend. MT is an amazingly flexible tool, but it is slow, slow, slow, and it’s not set up well to deal with comment spam. It takes several minutes just to delete a single comment–and then you have to rebuild all the static files–and I just don’t have time for that nonsense, sorry. I spent a month or so fiddling around with the MT-Blacklist plugin, but that’s slow too, and not very effective: Comment spam keeps using new words and new URLs, and MT-Blacklist can only filter stuff that you’ve already flagged. That meant I was soon spending an inordinate amount of time fiddling with MT-Blacklist … and as you can see I have hardly had time to post at all in the last month, as a result.
WordPress, by contrast, is fast, elegant, requires no rebuilding (since all of its pages are dynamically-generated PHP), and makes it incredibly easy to moderate and delete comments, or to turn them off entirely if you choose. I feel like a new man.
Bear with me while I play with templates for the next few days; I’m still trying to find the right look for this site (preferably without doing any actual CSS work myself; I just want to find a nice template and drop it in).
Dylan – Congrats on the move to WP – these migrations are a lot of work.
I think the question of whether MT’s “All the hard work up front and then no work for statically served pages” model or other DB publishing systems “a little bit of work on every page request” is interesting and inadequately researched. I’d love to know definitively at what level of traffic, and on what kind of server, one crosses the threshold from dynamic being OK to static generation being a requirement.
Not nec. an apologist for MT, but I think it’s basic philosophy here makes a lot of sense at some level. The biggest problem is comments of course — it’s one thing if the publishers needs 20 seconds to generate a page, another thing altogether if a comment takes that long. Hammer a server with 60 comment spams at a time and you have a real problem.
From your post, it sounds like you never got to play with MT3 and the new MT Blacklist, which make *tremendous* strides. By default, comments on posts older than 14 days or containing more than 5 URLs (configurable) are held for moderation. Which means they’re placed in the database but no pages are regenerated. Which means comment spammers don’t hammer servers, and despamming is similarly fast. The best part is that 95% of comment spam fits these criteria, so you catch the vast majority of it even before a URL is blacklisted. Things are *much* better than in the bad old days.
But WP is a great system (though missing a ton of plugins etc. that MT has) and I’m not knocking it. Just my .02.
P.S. – Can you make this comment entry box a few lines taller?
Thanks for your comments! In this case, the migration was pretty easy, actually. I agree that the dynamic vs. static page generation issue is debatable. In the case of the tweney report, it’s pretty clear cut: It’s a low traffic site, so the impact on your server is probably minimal, and I think my time is more valuable! But with a higher-traffic site (like Battelle’s, perhaps) the dynamic page generation may not work so well.
I never did play with MT 3.0 — I was turned off by the fact that I had to pay for it if I wanted to manage more than one blog. Plus I’d already been so irritated by the older version (even with MT-Blacklist) that I lost confidence. I feel like MT is the Microsoft of the blog world — they make great software, with tons of feature, but it’s also huge, bloated, and quite possibly riddled with security problems (I don’t know that for a fact, I’m just nervous about it because the code base is obviously monstrous). By contrast, WP seems lightweight, lithe, and manageable. Sure, it lacks plugins, but I never used many MT plugins. Right now, I feel that I need to focus more on writing my blog, and less on tweaking it. So WP suits me well for that.
Also, being written in PHP, WP may not have the need for plugins that MT has — it will be easier for many of us to extend the system on our own with basic PHP coding skills (although the key to the success of plugins is that there be a well-written API so that plugins are just that, rather than hacks to existing platform code that will get overwritten on the next upgrade.