The Great Change: Turning Cathy into a Lawyer

I recognize that some of my worst blog posts have been in the housecleaning category when I've gone twelve rounds with trying to upgrade my blogging software. There you've seen full-blown temper tantrums in response to the immense frustration such undertakings have necessarily caused. In the end I've always succeeded, but not without insignificant suffering.

The thing to realize in reading them, however, is that they don't reflect a lack of composure on my part nearly as much as they reflect a surfeit of defects on the software's part. In other words, yes, maybe it would have been better not to flail about so publicly, but such flailing was certainly not unwarranted. In fact, it may say numerous positive things about my composure that I only flailed about this much...

Nonetheless, criticism is always more weighty when given in a cool, calm, and collected manner, so I will endeavor to restrain myself before launching into the following complaint: MovableType is a complete pain in the ass to deal with.

I started using it nearly half a decade ago in large part because it was the only blogging software I'd ever heard of that could be installed on your own domain. Back then it was at version 2.something. Since then it has worked its way to version 4.01, via a major jump to version 3, lesser incremental jumps to version 3.3 or 3.4, and another major jump to version 4. I suppose the major jumps are a necessary evil; whenever you design a major software product you always want it to scale to new uses, but it's sometimes hard to anticipate them so far in the future. After a while it gets to the point where you can't continue to improve on what you've got - you have to start over from scratch.

As a MovableType user it's kind of distressing when this happens because it means you'll have to start over as well in order to upgrade. Hence my previous frustrations, as the upgrades often killed all my customizations and styling, and at times my posts themselves. Many people have chosen to avoid the frustrations by not upgrading at all - I know quite a few quality blogs that are still at version 2.x. But upgrading may be a necessary evil, as the newer versions close security holes while also promising new, exciting, and even useful capabilities the older versions were without.

It's this promise which is at the root of my discontent, because I so often find it so often undelivered. To be fair, eventually I came to like version 3. I could see how its general design offered some significant improvements over version 2, mostly by separating more effectively the content of the page from the design of the page. Of course, this separation meant that large parts of my site broke during the upgrade, but eventually things settled down and third-party developers started providing other tools and plugins to make using the MovableType product a relatively easy proposition.

Not content to rest on their laurels - which I suppose is reasonable - MovableType's developers decided to provide even more functionality. Which sounds good, but unfortunately doing so apparently required yet another complete reconceptualization of the product. This time, instead of better separating the look-and-feel elements of the blog from the content elements like they did in the last major upgrade, this time it was going to further split out the rendering logic from the page content as well.

Generally speaking, this approach seems to make sense. The power of the web in general is best realized when all its aspects (display, content, logic) are separately provided. A database holds the content, which middleware logic accesses and compiles into a page of output that a stylesheet then controls the display of. The less discrete each aspect is, the less flexible (and therefore the less powerful) each will be.

But after taking a look under the hood of the MT 4.01 version I recently installed I'm perturbed: I can't figure out what's going on. I see in general what it's trying to achieve, but exactly how it tries to go about it is a mystery. Which is extremely unfortunate now that I'm trying to work with it.

My needs themselves are modest. Initially all I'm looking to do is set up a site similar to the site I've already got. I have a second project lined up that might need to push the envelope a bit, but at first all I need is something simple: a nice website with blogging capability and a simple yet somewhat customized appearance. And I can't do it.

Well, let me rephrase: I'm sure I could do it, given enough time. But the fact that I need to employ skills I haven't honed since the peak of my previous technical career strikes me as something very wrong. If you need to have professional-caliber web development skills to be able to set up your own blog, all the promise of a world where people can truly self-publish through their own presses is lost. Yes, people can still blog, using the various services out there (e.g., Blogspot, TypePad, etc.). But there are definite advantages to owning your own autonomous outlet, and it really shouldn't be this hard to do so.

My criticism of MovableType is based in part on the sense that when revamping the product for version 4, the developers made some poor decisions when it came to designing its architecture. Various PR statements I keep seeing with respect to how wonderful the 4.1 version- still in beta - will be make me think the developers might themselves agree that it missed the mark. Of course, the touting of the upcoming beta just adds to my feeling of having been an idiot for upgrading to a product, MT 4.01, when it clearly was not ready for primetime. In fact, so unprepared was it that few there's a dearth of third-party applications that might make using it easier. All these developers seem to be waiting for the 4.1 version themselves before they waste their time on work that will likely break as soon as it comes out.

But these problems would be much less significant if, for once, there was actually good documentation. Over the years the MovableType developers have bragged about new MT versions having better and better documentation, but in my experience it's actually ended up worse and worse. And not just worse, but often objectively and unforgivably bad. For example, there's documentation on their website that apparently only applies to the 3.x version but bears no indication of such fact. The only way one would know is because all the 4.x documentation bears the notation that it's all still in draft form!!!

The other day I went to see what would be involved with trying to set up the new website I wanted to make. I'd gotten the hang of version 3.x and figured it would only take a few hours to set things up under version 4. But I was wrong. It won't take hours: it will take days, maybe even weeks if I wait for the 4.1 beta to get finally released, which will hopefully (although the way things have gone in the past, probably won't) address the things about it that are currently cumbersome and cryptic with version 4.01. The administration screens' user interface provides few clues about how it's all to work, and there's no supporting documentation to help guide me through figuring it out either. It's a terrible state of affairs, and if I'd paid money for this I'd be demanding a refund.

But that's the thing: I'm using MovableType for free. In fact, the company that develops it is further encouraging warm, fuzzy feelings with its recent announcement that MovableType will be open source. Huzzah! This is great news, as it's always good to see powerful software be offered to the public (both as users and further developers) on financially-accessible terms. In the long run this move will make MovableType a much better product and its developers much richer...

In the meantime, however, the current version is crap. And it's got me boxed into a corner. I really want to make those sites and I want make them now. But I can't. I certainly can't make them both expediently and also the way I'd like them to be, and I'm not even sure at the moment I could either achieve expediency or make the way I'd like them to be. As it is I think the best I could hope to do is to set something up with a basic stock template and style and then change it over time as further upgrades make it feasible. Unfortunately that means I'd have to commit a cardinal sin of web development by launching a website that's not actually done. While it's always fair game to improve and tweak a site, wholesale changes to a user interface should really be done before a site is launched and users become acclimated to it. Just as major software products should not be released before they are done and users (try to) become acclimated to it...

So frustrated have I been by all this that I even looked into chucking MT altogether and instead switching over to WordPress. I generally hear much less bitching and moaning about WordPress, and at first glance the administration of it does seem much less cryptic than MT ever was.

But the only viable solution I've ever seen for comment spam is a MovableType solution, and I don't think I could live without it. Plus, in the long run I wish to maintain several blogs on my website and take advantage of some of that promised functionality, so with MovableType I think I am stuck.

How that will affect anything you'll see here will eventually become apparent. In the meantime, however, I just needed to complain...