Second Update:I didn’t choose the Hubble article as a test case for my conversion because it’s particularly good writing. I wrote this years ago and I know it needs severe editing and updating. I picked it as an example primarily because it fit all the test cases for XHTML compliance.
Thanks to Jay Allen’s helpful notes in my comments, I found a simple way to generate an RDF/XML file for each of my postings, all using the PostCon vocabulary. You can see an example for this entry, in fact. Which means I can automatically feed these into my PostCon system for management. Which means…
Yeah!
(I really do need to take to time to read the MT documentation a bit more carefully than I have done.)
Additionally, as Jay pointed out, with the template/archive management, you can split category pages by directory (and rename the files to something other than the number system if you want), but I still prefer separate weblogs for major type of document, and to reserve the category for the specific domain. This gives me hierarchical categorization — type/domain (or domain/type if I had reversed it). Unless you can do this with Movable Type, too, and I didn’t see this functionality in the documentation, either.
I think a Movable Type book is in order, but I’m sure the Trotts are already writing it.
When taking a break from working on the book, I’m slowly moving my content over to Movable Type management. While doing so, I’m taking the opportunity to update everything to XHTML 1.0 strict, as well as being CSS2 compliant.
I’ve already made the transition to XHTML 1.0 strict in the main Burningbird Network page. This page lists out excerpts from new entries across all of the web sites, including this weblog. At this time, only the Practical RDF, Photo Gallery, Articles, and weblog links show in the page, but I’ll add the other sites as I bring them inline.
I pull the excerpt information from the MySql database (oh, how nice it is, too), using PHP. In addition, I created a Perl program that takes the RDF/RSS from all of my websites/weblogs and merges them into one RDF/RSS file. You can see the code for this in Chapter 14 of the Practical RDF book, which I uploaded last night. The application is scheduled as a cron job (scheduled re-occurring job) running every hour.
An advantage to pulling excerpt information for each posting for the main Burningbird Network page is this forces me to write an excerpt for each posting, article, tutorial, etc. Even if the excerpt is no more than the first few sentences of the posting copied over to the MT excerpt field, it’s better than the chopped up text that shows in the RSS feeds now. And it only takes a minute at most.
(BTW: If you’re subscribed to any of my RSS feeds, I would ask that you switch your subscription to the new file. However, you can also continue to subscribe to a specific entity, such as this weblog, or Practical RDF, or Articles, or whatever.)
I’m in the process now of setting up Burningbird Articles. In my past life, at each of my web sites I used to have a sub-directory of articles off of the main site containing more in-depth writings. I still do, but in addition to accessing the articles using http://yasd.com/articles or whatever was used, readers can now access articles at articles.burningbird.net. I’ll be doing the same to a all my other sub-directories — each one becomes a separate weblog.
I used a separate weblog specifically to control the location of the files — that’s the main determiner. Ultimately this becomes a content type category — such as article, tutorial, weblog, etc. Following this scheme, http://domainname.com/articles became articles.burningbird.net and is managed through Burningbird Articles; http://domainname.com/interaction became interaction.burningbird.net and is managed through Burningbird InterActZone and so on. I’ll convert all the older material I want to keep into weblog entries, and since the file names will differ — MT does use a numbered system and my old entries had regular names — my post-content management system (PostCon) handles the re-directs to the new locations.
All main pages are PHP and full of PHP goodness. All secondary pages (individual articles, posts, etc.) are HTML pages — no server-side coding. Peripheral pages such as Backtrack are also PHP or based on some other functionality such as CGI.
Another significant change is my domain management: I used to have several sites, each with their own domain: YASD (yasd.com) for computer technology; P2P Smoke (p2psmoke.org) for distributed and P2P technologies; Dynamic Earth (dynamicearth.com) for science and environment; Solar Lily (solarlily.com) for art; and NetJetter (netjetter.com) for travel, hiking, outdoor recreation, and adventure. I still have the domains, but all of the sites are now merged into Burningbird, and all point to burningbird.net. They’re joined by a couple of new ones: MirrorSelf, which will become my photo blog in addition to my static photo gallery; and EvilWoman, which is going to become something….else. My online book, which I’ve been working on quietly offline, is called Marbles, and it will be at marbles.burningbird.net, when I set it up.
To maintain the distinctive flavor of each item as it moves over to the new system, I’ve created MT categories in each of my site weblogs (such as Burningbird Articles) and labeled then ‘yasd’, ‘dynamicearth’, ‘solarlily’ and so on. When an individual page is opened (such as my Hubble story I just ported) I’m using the category to change the style sheet as well as logo, using the MT tag <$MTEntryCategory> tag. You can see it in action now — all main pages willl eventually open with the Burningbird deep burgandy color (the bittersweet chocolate brown color is going away) but anything associated with Dynamic Earth, such as the Hubble story, opens with a deep pine green. P2P smoke is gray, NetJetter is a deep blue and so on.
As for layout, all main pages have the three column layout shown in the Burningbird Network web page and the primary Burningbird Articles page; all secondary pages have the two column layout, as shown in the Hubble article page.
The only exceptions to this rule will be this weblog, the Practical RDF weblog, and Marbles.
The layouts not only validate as XHTML 1.0 strict, they also conform to Mark Pilgrim’s Accessibility guidelines, as you’ll be able to test for yourself if you use a voice browser (just the main page for now).
The clean up for each article is enormous, especially considering my use of graphics. I had replace the IMG vspace and hspace attributes with CSS margin attributes, and the align attribute with the CSS float. The result is better, just a lot of work. Not to mention closing all of those open break tags. And why did I use caps so much for so many of my tags?
Another change I had to make was to remove the ‘&’ from all CGI references int he MT template code (such as with trackback and setting the view mode). I had to replace them with the encoded value, which I’m not sure how to show you.
However, in spite of all the care, I am still having problems with IE. No version of IE will take follow the CSS height attribute when used in a TD element. Based on this IE oversized some of the rows, and the looks are off. No problem with the writing, just the overall look. But I’m not going to change the design because of a flaw in Microsoft’s IE browser.
(If someone has a workaround, I’d be glad to hear it.)
One other change I’m making isn’t ready to view yet. I’ll post when the application to support it is finished.
If I had one wish for Movable Type, it would be the ability to create a separate template that’s anchored to an individual entry so that I can create a separate XML page associated with my individual archive entries (other than the XHTML ones). However, MySql comes to the rescue again — what I can’t do within Moveable Type directly, I can do within the Movable Type database.