I’m taking advantage of this server move to make some pretty drastic changes in my own sites. For instance, I’m not going to try maintaining the old numbered system for my Movable Type page names because, to be blunt, it’s a mess.
What with my recent tax evasion weeding out, and my habit of splitting weblog entries across different sites, the numbering is completely out of whack. Enough to bother even me, virtual slob that I am; for the anal among you, it would be enough to send you into a coma.
I contemplated a weblog redesign — something all new. My first thought was to put a big graphic at the top of a half naked man, but then I thought, what does this tell people about my weblog? So I discarded that idea. The For Poets sites has a look I like and I considered using it with this weblog, but displaying different photos every time you access the page. However, this idea is too much like Jonathon Delacour’s and I don’t want to steal his mojo.
Besides, I like my weblog look. I’m used to it. It suits me and what I write about. I may, however, change the look of the photo blogs, and I’m definitely changing the rest of the sites, such as burningbird.net.
The photo blogs are going to their own domain, mirrorself.com. All my photographs are going to this domain, and you can imagine how interesting this is going to be with all the embedded photos I have in my pages, and the number of photo blogs I have (each with hardcoded absolute URLSs). The For Poets weblogs are also going to their own Movable Type installation, and will be using the new page naming system. There aren’t that many For Poets weblog entries so doing redirects could be handled manually. However, with Burningbird and the rest of my stuff, we’re talking a significant impact. I have an application that’s currently tracking requests for missing resources and all I can say is, you sure can tell I’ve been online a long, long time, and that I move things around a lot.
One challenge with splitting my weblogs into completely different MT installations is my current comment/trackback facility. Normally this goes across all the weblogs; through this approach, to be blunt, I own Blogdex, as a comment for one post is repeated across all weblogs and robots see this as a fresh link to the post. I’ve been in the top Blogdex ranks every weekend for two months (weekends are slower linking times.) I’m trying to decide if I’ll find a way to work across databases, or to be kind to Blogdex.
To handle the Burningbird weblog reorganization, I’m putting my little PostCon application into full gear. The only part missing on the application is the forms-based front end that allows you to create a PostCon RDF file from scratch. I really don’t like doing forms-based development — I like working backend stuff. However, I don’t need to have the forms-based component right now. It would be handy, but I don’t need it.
(What would be nice is a generic forms application that can be used to define a data model, automatically create the forms, and then record data to create the serialized RDF/XML files. Wait a sec, I do! It’s called Protege. I’m using Protege for my PostCon pages that aren’t being generated through Movable Type.)
I integrated PostCon into Movable Type some time ago, but now I’m increasing the integration and am using pieces of PostCon, as well as Movable Type to handle the redirects — from old numbered pages to the new page system. More than that, though, is that each page now has its own particular history — what did the resource used to be named, what is it now, who wrote it, what’s it about, and linkage info. All in a machine readable format, that can also be viewed by people pushing a button on each individual page and seeing the ‘hidden’ page self-description. There’s a little FOAF in this, as well as a few other odds and ends RDF vocabularies that I’m absorbing.
I’ll be writing all this up in my Semantic Web for Poets site. I hope to show that the semantic web starts small, and starts when each of us takes a little bit of extra time to record just a little bit of extra information that could be helpful down the road. Yes, PostCon uses RDF. But it also uses plain old, Perl, too, and is served through Apache, and run on Linux. The entire Internet did not have to be rewired in order to use it.
For those who like moving parts, yes, there’s even some moving parts, though my weblog still doesn’t talk to my toaster.
Caveat on all of this, though: There is going to be some major changes and expect a rough week for my sites. Not for anyone else — the other weblogs should move with a minimum of fuss and bother.