Author: Barry Parr

It looks like I have to add a category for full CMS’s with a sufficient layer of abstraction to build sites other than a blog, wiki, or Slashdot clone. The three that have been mentioned in comments are:

EZPublish is a PHP/MySQL CMS that looks like it does all the right things, and you can get it as an abstract framework, or preconfigured for more structured sites. However, the one (undisputed) review that I have been able to find says that EZPublish is crufty and slow.

Bricolage is aparently the open-source evolution of Salon’s CMS, using Perl and MySQL. Personally, I’m leery of Perl, which I don’t know, and PostgreSQL, which I don’t use and can’t pronounce. But it looks like a mature and exciting alternative[too-short eWeek review].

Typo3 is abstract and uses PHP and MySQL. However, I’m having a really hard time understanding it by reading the site, which is badly organized and poorly documented. The site itself says it takes weeks to learn how to use Typo3.

If you’re looking for a CMS with a greater level of abstraction, you’re going to need to spend more time on your decision. I’d probably install both Bricolage and Typo3 and get to know them both.

Perhaps newspapers will never understand the Web.

We’re approaching the ten-year anniversary of newspapers on the Web — Mercury Center (my site) and SFGate launched at the end of 1994. Our vision of the Web has changed a couple of times in the last decade, but newspapers vision of their online edition remain unchanged.

Right now, we’re in the middle of a bottom-up revolution in how the Web is created and how people use it. OK, all real revolutions are bottom-up. That’s how you know it’s in a revolution.

Right now, Technorati is indexing 2.5 million blogs. Most of those are inactive, and most of the rest suck. But there is a huge, unmanageable number of sites remaining that are changing the way that people use the web. And the tools they use to create their sites and reach their audiences are steadily improving.

The title of this piece is deliberately provocative. I don’t expect newspapers to mimic blogs. But I don’t understand why they haven’t learned some broader lessons about how our use of the Web has changed in the ten years since we first went online.

Newspapers are treating RSS as a threat to their core business. They are desperately afraid of “aggregators” grabbing their headlines and treating them as wire services.

Why are they afraid of aggregators? I understand the rationale, but it doesn’t really make any sense. They want you to visit their home page, which they view as the gateway to the rest of their site, every day, whether they have any news for you or not.

Publishers don’t understand that the home page is no longer the gateway to their site. Every well-designed page has enough navigation and headlines to draw you into the site.

Publishers don’t trust their newsrooms to deliver headlines that will bring you to their site because you have to read the story.

Publishers are anxious because they can no longer get you to pay to have them deliver a package on your doorstep that you feel compelled to read because you paid for it and because you’d feel guilty to toss it out unread.

Publishers want you to read their sites because it’s a habit and not because they’re producing must-read journalism.

I spend too much time thinking about cheap content management. Between new sites and new licenses for software I’m already using, I’ve got a couple of reasons. But I think I may just be compulsively fascinated with the idea that my ideal CMS is just around the corner…and that it’s free.
My criteria are simple. A CMS needs to be able to create sites more complex than a simple blog. It needs to be easy to install and use. The software must be mature and apparently bug-free in daily operation. It should have a large community of users and developers. And it should cost less than $200 for a commercial license.
There seem to be four principal differentiating factors among these CMS’s:

• Platform: Perl, php, or Python/Zope. If this matters to you, you already know.
• Page generation: Static or dynamic. This is somewhat platform-related. For example any php site is going to be dynamic.
• License: Commercial, or Open Source. I’m not a zealot about this. I love the idea of Open Source, but I’m currently using commercial packages for my sites.
• Type of site: Weblog, news site, or Wiki. What you choose depends on what you’re publishing. How chronological do you want to be? Do you want a lot of modules packaged with your software? Do you want your site to look like Slashdot, or do you have an original design in mind?

So, here are what I think are the top ten free and cheap content management systems, in alphabetical order. If you’re thinking about creating a site, this would be a good list of candidates to start with. Examining each would also help you work out your ideas about the ideal CMS for your application. I’ve included a couple of basic blog packages that might not meet my personal criteria, but which I know people are using creatively.
Blogger is free and you don’t have to install any software. If you don’t know why you’re blogging yet, this might be a really good place to start. It’s owned by Google, which is a plus or minus, depending on your point of view.
Drupal is a full-blown site management system (php and MySQL) that has gotten a lot of recommendations since Movable Type changed their license. It’s open source and based on php and MySQL. It’s part of a geek triumvirate with Plone and Slash and I’m wondering if I really need three packages in this category. Other php-based CMS’s in this category are Nucleus and phpNuke.
Expression Engine is the newest CMS from Rick Ellis, who created pMachine, which I use to run Coastsider. It’s based on php and MySQL and seems very powerful and flexible. It will cost you money to run, either for personal or profession use, but it’s inexpensive and the license is flexible.
Movable Type is what I use to run MediaSavvy. I love MovableType. It’s based on Perl, but I love it anyway. It has a huge user and developer community. Movable Type pages are not dynamic and have to be rebuilt every time you make a change. This is reasonably fast, but can be a real pain if you get a lot of spam comments. They have the best templates in the industry and an inordinate market share among A-list bloggers.
Plone is based on Python and uses the Zope platform. It feels sort of like Drupal and Slash, and they’re all designed to help geeks reproduce Slashdot in whatever realm they’re geeky about.
Slash is a perl-based system for Slashdot-like sites. You need to have root access to the server it runs on, so it isn’t going to work for most users. A similar package with the same limitations is Scoop. I seriously considered using Scoop, but it was missing a lot of the things most modern CMS’s should offer, such as real templates and CSS support. I don’t know about Slash.
Some Wiki or other, there are dozens, I can’t tell them apart, and they all make my head hurt. But wikis are undeniably cool, ideal for some applications, becoming a lot more sophisticated, and are beginning to look like an overnight success ten years in the making.
WordPress is another contender who’s profile has been boosted by Movable Type’s licensing misstep. It’s php and MySQL, and its open-source. And it has the momentum of a killer asteroid. It’s biggest limitation is that it can only handle one blog, so you need multiple installations for complex sites. However, at least one thoughtful fellow chose it for his complex site.
That’s about ten, depending on how you count. I’d be interested in more nominations if they’re serious contenders for top ten and genuinely different from the ones I’ve listed here.

I love Movable Type. Movable Type changed my life. Movable Type is a better piece of software for me than anything I could have designed for myself. I would pay money for Movable Type. I’d like to see Ben and Mena Trott get rich from Movable Type. And I’ve been looking forward to the release of version 3.0, while they’ve been focused on other projects.

But their new licensing scheme is confusing. I think Inluminent hit the nail on the head: they need an unlimited personal license for a reasonable flat fee for personal use (say, $100 for unlimited users and weblogs) and either a flat fee or user/weblog licensing scheme for commercial users. Personal users shouldn’t be subjected to tiered licenses.

I don’t think anyone who is a serious enough blogger to need a lot of users or weblogs from Movable Type would object to paying $100 for such a wonderful tool.

But right now, I’m not sure what I’m going to do next.

I’m tempted to move to Expression Engine for its simpler license, dynamic php publishing, and the free “switcher” license they just offered me. I’m using their pMachine software for Coastsider and am considering migrating that site to Expression Engine.

But MT is making me consider whether I want to use commercial software at all.

I have been so happy with the GPL software I’ve been using lately that I’m tempted to move all my personal publishing to GPL as a gesture of support. WordPress seems to be the GPL solution of choice, with a large community and lots of momentum. It also uses php, which is quickly becoming my platform of choice. But it doesn’t have the author and weblog management flexibility of either commercial option.

MT 2.x does what I need right now. Until I need the features MT 3.x or Expression Engine, or WordPress matures, I think I’ll sit tight.

Gerry McGovern says we should treat content management as a process, not a project. One example of this kind of thinking is making sure that the things we build are maintainable once they are built and that a procedure for maintaining them is in place.

Gerry’s talking mostly about content management on intranets, but this is clearly applicable to online publishing.

I remember a meeting in the early days of the Web when our project manager told the team, “When the server upgrade is finished, things are going to be a lot better.” To this our webmaster replied, “Don’t you see? We’re never going to be ‘finished'”. I’ve told this story over and over because it points out the difference between two schools of thought in IT generally and the Web in particular. I wouldn’t call our webmaster a pessimist, although the project manager probably perceived him that way.

I’ve always readily admitted that it’s a lot more fun to build Web sites than it is to maintain them. If you’re not careful, this kind of thinking can lead to the creation of lots of cool features that don’t work all that well and eventually break. Gerry says that this point of view also leads to the idea that everything will be fixed in the next redesign. But, if you don’t change your way of thinking, you’ll find yourself in the same fix when you’re done.

Gerry says we should be building a small number of important features (e.g. search, directories) that exceed our users expectations and can be maintained in that state over the long run.

I really like Rafat Ali’s new Digital Media Jobs Blog. It’s an incredibly simple idea with a real revenue potential.

He’s simply blogging digital media jobs, from the usual online classified sources. Finding these jobs in the usual haystacks is nontrivial and the number and quality of his listings is high. This site demonstrates how versatile the blog format can be and that we have only begun to explore its possibilities.

After spending too much time an energy dealing with posting spam, I finally installed MT-Blacklist. I’ve been concerned that it wasn’t going to be effective, but it was becoming clear that the vast majority of the spam in my comments was coming from just a couple of scumbags and that a blacklist was probably the right approach.

Comment spam has become such a nuisance that every time I open my mailbox I could feel the tension and anger rising.

While I don’t have enough experience to say how well it works, it did manage the few test messages I threw at it. It is really well-designed and easy to install and set up. It also allows me to avoid inconveniencing my users by requiring registrations for comments or closing comments after a certain period of time. Finally, it avoids the more radical and time-consuming solution of switching my site from Movable Type.

Yesterday, I wrote that I received an obvious spam comment, apparently from Edmunds.com. Today, I was able to analyze my log files, and the results are explosive.

At the date and time that message 364 was spammed, that message was accessed by a Java program from IP address 66.161.49.254. According to the following IP Address lookup, this address belongs to Automotive.com, a competitor to Edmunds.

Automotive.com also spammed my site on 3/10. Finally, I was spammed via email, apparently by one of their affiliates using a forged return address, eight times in November and December of 2003.

Edmunds general counsel posted a prompt reply to my posting yesterday evening, denying they had anything to do with it. Now that we know who is responsible, I encourage him to take it up with Automotive.com.

Update: Edmunds didn’t spam me. Automotive.com spammed me, using Edmunds name, and I have the log files to prove it. Also see the comment on this message from Edmunds general counsel.

My web site has recently been plagued recently by spammers trying to scam links to sites offering viagra and baldness cures.

Today, I find that someone calling himself Mark Riley has posted a completely irrelevant and gratuitous comment on my site with a link back to edmunds.com. Here is the comment. You decide if it’s anything other than comment spam.

Edmunds.com is a legitimate site and one that I use and like. Are they now using spam to promote themselves? I can see no other explanation for this message. No one other than Edmunds could possibly benefit from it. I’m awaiting Edmunds’ reply.