I have been committed to Movable Type, but I have had some problems with it. When friends like Randy with history and good connections in blogging talked about changing to WordPress, I tried to compare the products. I think SixApart and MT will lose ground if SixApart doesn’t address some problems. I don’t want to spend time on importing entries and writing new stylesheets and templates for a WordPress blog. For the time being, the balance of convenience favours staying with MT and hoping for improvement.
SixApart has been beta-testing a new version of Movable Type, MT 3.2 for about a month. They will be releasing the 4th beta version of 3.2 tomorrow, trying to find and work out the last bugs before rolling out the product. Since uploading MT files takes more than a few minutes, I prefer to have a stable product on the first release. Possibly, the new release will deal with some of my concerns.
WordPress fans claim that it has advantages, and I haven’t seen MT trying to respond, either with information about their own product’s features, or announcements of product upgrades. It may be that they are just not spending the time to answer unfounded criticism and rumours, but they do not seem to spend time and energy on documentation.
WordPress claims simplicity of installation. I think it’s safe to say that MT installation is intimidating for anyone who does not have experience with scripts and PHP. MT’s documentation was written by geeks for nerds. It assumes technical knowledge that many users don’t have, and it is not clear. This points to MT’s greatest weakness and its largest marketing and PR problem. A few years ago it was open-source, free, and a nerd’s delight. When MT went commercial, the perception was that MT had evolved into a deluxe tool for online journalism and publishing. But the corporate culture of 6A still seems to assume that the customers are technically savvy, able to understand code and willing to enter lines in their mt.cgi files, and write code and tags in their templates and text. It isn’t easy to install or to customize. Meanwhile the nerds are migrating en masse to WordPress because it is the current cool open-source product, and more pleasing example of the programmer’s art – clean script, XHTML compliant.
MT and WordPress are both content management systems designed to publish text on a dedicated server. Both systems allow a use who knows something about HTML, CSS and PHP to customize the appearance of the published blog. Both systems come with default templates and stylesheet, and templates and stylesheets (WordPress has themes) are easily imported and installed. I can’t really say which system has the best options. Older MT Styles and templates do not feature optimal HTML. Beyond that, you have to go into your stylesheets and templates and set colours and tags.
Blogging with hosted systems like SixApart’s Typepad and Livejournal avoids a lot of the wear and tear of design and maintenance. They have their own design interfaces which let a use change the look of the blog, within a range of options. Elise Bauer’s Learning Movable Type blog wrote that a Typepad blog can do what most users want, if you don’t mind being contracted to a hosting service, rather than a server host, and if the hosting service gives you the right features and options. My daughter Claire started with UJournal, a subsidiary of Livejournal, and has continued blogging with Livejournal. The Livejournal service has interactive features, beyond content publication. The hosted services provide some features that aren’t part of a content publishing system, but they are also limited in the features offered, and I don’t think features can be added. Typepad and Livejournal can download entries, but Livejournal won’t import except by cut and paste, entry by entry. That particular feature may be important if you migrate to a different system. MT and WordPress both handle a variety of plugins support customized features.
I really don’t know if WordPress would be easier to handle in design and maintenance. Perhaps its cleaner code has some advantages. I have seen flaws in MT’s manuals and support literature. I haven’t gone into the WordPress literature to see if it is any better.
WordPress publishes all entries dynamically. MT’s default is static publishing, which is not really a problem in my opinion. It is a little annoying to wait for pages to rebuild when new entries are published and when templates are changed, but it doesn’t take more than a few seconds. Dynamic publishing has been an option in MT since the fall of 2004, but there are problems. There is an installation process involving writing an htaccess file, and configuring templates. MT itself cautions that the server load of dynamic publishing can be a problem for some blogs. I don’t know if this is a bigger problem for MT than WordPress, and I don’t know what kind of demand may cause a problem. The third problem is that MT plugins usually don’t, or didn’t, work with Dynamic Publishing. MT-friendly sites like Elise Bauer’s Learning Movable Type have articles warning that most MT plugins were or are scripted in Perl, and won’t work with Dynamic Publishing. This information has not been updated, and the articles link to a tech article writing about porting to PHP or writing for Smarty. I would hope many of the plugins have since been made compliant with both systems – but I don’t know. MT’s plugin directory does not indicate which plugins work with Dynamic Publishing and which ones don’t.
WordPress (and some hosted services like LiveJournal) allows a user to password protect an individual entry. This is a useful feature for someone who may occasionally post an entry for the information of a few friends and visitors on an open and public page. This feature has arrived in MT by way of a plugin in the last few months but it isn’t clear the plugin works with Dynamic Publishing, and if it will continue to work with MT 3.2. It apparently doesn’t work with the 3.2 betas.
Plugins are freeware, and the designers have no reason, except personal pride, to document and support them. MT isn’t supporting them except by linking in the Plugins Directory. Some of the functionality of MT has depended on plugins. For instance, all the useful comment and trackback spam control came from plugins. MT has been building some of the more useful plugins into new releases. In a previous release, it built in sub-category support, and the new release has an improved category interface. MT 3.2 will include Spamlookup as part of the default installation. Still, MT could do a better job of explaining these issues and helping non-tech users to use the tool as a whole.
Wordpress only supports one blog. For some users, MT has an edge because it runs multiple blogs off of one database, allowing system wide spam settings, and system wide installation of plugins and upgrades.
MT has some editing quirks. I have had a problem with formatting text in quotations. It is a known issue. When MT converts carriage returns, some unexpected and annoying things can happen to the published page – nothing deadly, but extra spaces, odd indents, unwanted line spacing can appear. The workaround involves extra care in writing the entry text to avoid creating invisible markup tags. Is this acceptable performance in a commercial product?
If MT could fix some of these annoyances …
As I was writing this post, I ran into a problem. I could not save my post while it included the word Perl. That word used in a sentence, followed by a trailing space, led to an Apache error message. I have worked around that, but that’s a nuisance. [Update] The problem was with Blogomania, my server host, which had implemented security measures to defeat cgi scripts.