I started my third (and likely last) weblog sometime around January of 2006 (I almost said earlier this year... but it's 2007 already) and because it was going to be more of an experiment in writing (i.e. seeing if I would perhaps consistently write about things that I felt really didn't fit the purpose of this or my other weblog), I figured I'd also see if I wanted to move to another/different weblog software system. I had already had some experiences with Drupal and it left me with a bad, bad taste in my mouth. It came down to the fact that there was no coherent way of doing anything if you were coming to Drupal from the perspective of MovableType. I say this because it is possible that my perception of how poorly Drupal functioned had to do with my previous experiences with CMS (content management systems) like MovableType. Rather than having a somewhat coherent set of documentation to describe how the system is intended to function, one had to rely upon user/developer provided documentation. Usually, said documentation was lacking in function - i.e. a technical document should have a purpose and how section, at the very least, and consistency among different sections. Drupal's documentation didn't. For example, the term taxonomy was used to describe a user authentication and rights system. Taxonomy? Out of what orifice was that term removed? (According to OS X's built in dictionary, taxonomy means: the branch of science concerned with classification, esp. of organisms; systematics. In my head, this relates poorly to authorization and user rights.) Perhaps this is a poor example. I am trying to give Drupal the benefit of the doubt and it may be that many people will find it a usable and reliable CMS. I didn't, however, and that is the point of the above comments. I found the documentation confusing and incoherent when viewed as a whole. This is not to say the software itself was bad, but my opinion is that software developers need to spend as much time, if not more, writing an explanation of their code as they do coding.
So, with that experience in mind, I looked at the other alternative: WordPress. Everyone seemed to be using it. Two of the bloggers in my top ten moved to WordPress sometime in the year prior to me testing it; Jeff Jarvis of BuzzMachine fame and Erik Barzeski of NSlog();. There had been general rumblings in the MovableType community regarding two issues: licensing and comment/trackback spam. SixApart had made the move years ago to monetize on the MovableType fame (and rightly so, in my opinion), and in the process irritated many (I'll never forget when they first changed the licensing terms and started [gasp] charging for the usage of MoveableType under certain conditions - especially where MovableType was being used in a commercial environment). I think the worst thing, though, they did, was let issues with comment/trackback spam languish. I know Jay Allen did a lot of work early on to deal with comment spam. This was before he was actually hired by SixApart. I stopped following his weblog after that (frankly, I don't remember why). I think I stopped reading weblogs for some time and his was one that I forgot to pick up again. The problem had gotten pretty big, though, and WordPress had seemed* to have made serious inroads in this area.
One other area of contention - one that I think most hobby users never cared about - was the load that MovableType put on a server as pages were re-indexed and as the site was essentially rebuilt every time a new post was made. This has never been something I really cared about. Most, if not all, of my personal web related stuff is on servers that I either own directly or control and manage. If I needed more horsepower to drive my own drivel, then I had yet another excuse to buy more hardware. Aaron has told me (I consider him a professional in the area of commercial weblogging), though, there are ways around this problem. Either way, it was one of the complaints, and because it was part of my decision to move and try another system, I'm writing it here.
So, with It's you, not me..., I made the jump to WordPress. What a mistake. Why? Simple (for me): MovableType, by default, creates real pages whereas WordPress's default mode (I'm not sure if there are other ways of doing this) is to create pages dynamically. For example, if I set WordPress to create daily and individual indexes of all my posts, you couldn't simply go to the directory in which WordPress resided and find those pages. I learned this the hard way when I couldn't figure out why WordPress wasn't publishing two posts I had just made (and yet I had no errors). After some head scratching, I finally realized the date/time stamp was wrong on my machine (two plus hours behind the real time) and because the posts were made at ca. 18:00 and the server's time was at 15:47, WordPress didn't show those posts until the server showed 18:00. At the same time, I realized that although I had set up WordPress to create individual and daily index pages, they weren't "real" - in the sense there weren't any files sitting there to be had in case the CMS died.
The last nail in the coffin for WordPress was something I wanted on a whim: export all of the content to MovableType. I didn't think this was that big of a deal since it seemed that exporting MovableType content to any CMS I had come across had already been developed. No deal. There is a plugin out there called WordPressExport, but I was never able to get it to work correctly (and believe me, I tried) and it is really too crude to really term as a semi-featured export tool.
This bothered me, not because I wanted to move to MovableType at that point in time, but because I simply wanted to make sure I had a backup of what I had written in a form that could be used in another popular CMS.
As I had experienced with Drupal, I also found the documentation for WordPress to be relatively inconsistent. When I had questions, many people (and Google) pointed me to the forums. I have a problem with this. For the very reason why the first source I hit for information on a given tool in OpenBSD are the man pages or the FAQ, I want the same thing for a CMS into which I'm pouring tons of energy and placing trust to have the same resource. I want an authoritative source for answers to basic questions. Forums and HOWTO's are far from this. (As a perfect example, look at the HOWTO I translated years back for qmail. I haven't updated that information in well over a year. I do have a disclaimer at the top of the page, but the information is still out of date.)
Sure, I believe in the power of user generated and open source software. At the same time, I believe that, for my purposes, commercially written software can be better suited to my needs. In this case, MovableType won out for two reasons: documentation and consistency. I simply like MovableType's tag system and don't mind the issues with rebuilding. I imagine there are many others, if they are honest with themselves, who will come around as I did. Two weeks ago, I spent two hours re-posting all of the content I had generated in WordPress over to MovableType at iynm.net. It has turned out to be well worth the effort. Within no time, I had a Flickr plugin installed and working and no longer have problems posting images inline.
MovableType is far from perfect. I still have issues with comment/trackback spam, but thanks to a [laugh] plugin originally for WordPress (for MovableType: MT-Akismet), that issue is largely resolved. I'm curious to see what other WordPress users will do. I already hear (and saw firsthand) about problems with sites requiring load-balancing and/or to be distributed across multiple servers. This is something I have no clue how one would solve in WordPress. It isn't, perhaps, a common problem, but it is one that seems to be caused by the way in which WordPress insists on generating truly dynamic content. Who knows what others will do in the end. I'm happy to be back with MovableType.
* I know I am using a lot of words like seemed, appeared, etc. I am purposefully softening my commentary because much, if not all, is my opinion and is based upon my subjective impression of the ongoings in the blogging software community.
Comments (4)
Eric Barzinski?
At any rate, you can't really have it both ways: you can't have "actual" files without rebuild times (though MT's still seem 10x slower than they need to be) and you can't have a dynamic site with "actual files." If you don't like hitting the database frequently, use WP-Cache.
Someone could write an WP->MT export tool for WordPress in a few minutes. WHy hasn't it been done? Perhaps because nobody has moved from WordPress to MT! I'm being slightly sarcastic... but not much.
Posted by Erik J. Barzeski | January 7, 2007 10:05 AM
Posted on January 7, 2007 10:05
Sorry - really, really sorry for the name spelling mistake, Erik. I've been reading your weblog for so long, and I kinda' knew in the back of my head that the spelling was going to be off. I was on a plane when I wrote this, though, and forgot to edit/check it when I ended up finally posting it (shame on me).
Anyway, on to your comment. My post had to do with the fact that with MovableType, I have my cake and eat it, too (for my purposes). My site is dynamic only in the sense that my main page and index pages are changing automatically as I add new content. I want static pages, however, for actually serving that content. I don't understand WordPress's methodology of generating pages (from a programmatic and, I guess one could say, philosphical angle) and I admit, I haven't taken the time to learn.
Yes, WP-Cache solves that problem, but for me, MovableType's way of doing things simply made more sense.
Moreover, I don't understand what the issue is with time to rebuild. I think I would have to agree that MovableType rebuilds are slow, but at the same time, I've never had an issue with that - and I still don't see why others would. If you were generating stories by the minute, this would obviously be an issue, but I'm not, so it isn't.
Also, as may have been implied in my post and my comment on needing more power as my excuse for more equipment, I don't care how often WordPress hit the database to generate pages. The fact is that I didn't like WordPress's out-of-the-box experience in relation to how I use a weblog and how I intended to use the software.
After re-reading what I wrote, I want to make one other thing clear: I don't mean to imply that Erik or Jeff will change back to MovableType or that they should. I stated some of the reasons why I did and am simply curious to see what happens with others. I make the comment:
"I imagine there are many others, if they are honest with themselves, who will come around as I did."
I only meant to say that I am truly curious to see if others will find the same issues I did. Perhaps so and perhaps not.
Also, you may be right in your implication that no one is interested in writing WP -> MT export plugin. Again, for me, the issue was that when I did try to export content, it was a pain in the ass because there really weren't the tools I needed to do so. With MovableType, there are - and if the reason is because people are leaving in droves, so be it. I'm simply happy that the tools are there. (Plus, as a side note, because I have more time and experience with MovableType, I don't find the process of going about doing so nearly as daunting as I did trying to export out of WP.)
Posted by Steven N. Fettig | January 7, 2007 8:26 PM
Posted on January 7, 2007 20:26
To each his own. If you're comfortable with your use of MT, more power to you.
But since you're sharing your experiences, let me tell you a bit more about mine.
With 3,000 posts and 15,000 comments, a "rebuild" even to fix a typo in some HTML took 45 minutes. Rebuilding a recent post - which updated the archives, the home page, the page itself, the pages before and after, and probably a few other pages as well - took about a minute.
Though I post frequently, it's not close to a rate described by "stories by the minute." It's simply a matter of accumulating a lot of data over time.
WordPress "works" the same way MovableType does, except that when MovableType "builds" a file, it then writes it to the disk. WordPress (without WP-Cache) builds the file the same way… then rebuilds it when the next person asks. A change made to fix a typo - or to address a massive change in my site's appearance - happens instantly, site-wide.
In the end, I could have almost tolerated the incredibly slow "rebuild times" of MovableType if it weren't for the incredible slowness MT had when spammers would hit my blog. I'd often see the server spike to - I kid you not - loads of 35 or 50! One time it reached 85! WordPress has been kinder and gentler to my server - with or without WP-Cache - and has a number of other perks - including a bevy of plugins - that made the switch worthwhile for me.
Again, to each his own, and may they enjoy the software they use. Because, let's face it - if we don't enjoy blogging, we won't do it quite so much.
P.S. MovableType's backup mechanism is lacking in and of itself. Perhaps one of the main reasons many platforms support MT import is simply because, three years ago, MT was the juggernaut. The fact that the export format hasn't changed a lick since then should tell you something, too.
Posted by Erik J. Barzeski | January 7, 2007 11:07 PM
Posted on January 7, 2007 23:07
See, this is why I missed having comments enabled on this site...
Erik, you make a ton of good points above. (And, in my opinion, those comments are as valuable as the original post.) These are all issues I have not had to deal with because, although, this site has been around for a long time (longer than my archives - since 1997/8) and has been running MT since 2003, I don't have the sheer number of posts or comments that Erik has.
I agree that 45 min. to rebuild is asinine. The spam load issue is also bs. It is one of the reasons I disabled comments completely for a while.
Luckily, I haven't had to deal with the rebuild-time issues - yet. Perhaps we'll see me do a complete flip and I'll be writing a post about how wrong I was. Right now, it won't happen because I'm done tinkering (other than css stuff) for the time being.
For the reasons I thought were emphasized near the end of my post, I still enjoy working with MT. I like the tags and I like the documentation. I am likely being unfair to the other systems out there, but, as you said, to each his own.
Thanks for the comments, Erik, and whomever else wants to chime in, please do...
Posted by Steven N. Fettig | January 8, 2007 7:17 AM
Posted on January 8, 2007 07:17