The Redesign and Relaunch of Molrak.com
Long time followers of this site (all 2-3 of you) will have noted a brief outage in the site around the new year, followed by a new look for Molrak.com, and since then a relative silence. This is not without cause. One of my New Year’s goals was to properly maintain my public blog here on molrak.com, which I’ve been neglecting for at least 2 years. When I sat down on December 30th to work on the redesign (I wasn’t really planning ahead as far as New Year’s goals go, I know), I first tried to identify why I hadn’t updated the old one much, and then I tried to come up with a few goals for what the new site should look like and do. The problems were easily identified, but the solutions were not so obvious.
I was not satisfied with the old design of the site, partly due to how it was hooked into k2. While the k2 theme for WordPress has many powerful features, it was another layer of compatibility which I had to struggle with. Instead of doing yet another k2 redesign, I opted to try to design my own derivative theme, using the default WordPress theme as a base (the default theme is based on Kubrick, which is the predecessor of K2). The default WordPress theme is incredibly well thought out, although I prefer to use %’s, ex’s, and em’s for my units instead of pixels.
I knew I wanted a redesign. I had a starting point, now all I needed was a destination. When beginning any design project, I start with either an outline or a thumbnail. For this project, the thumbnail came first. I had a vague idea of what I wanted the site to look like. For whatever reason I prefer the black background for my personal sites. Perhaps it’s because I still spend most of my time in front of a CRT monitor, maybe it’s my background in computers going back to Apple ][’s and the old Tandy 1000HX running DOS 2, or because I agree with Maddox’s view on staring into lightbulbs. I set my palette for dark swatch, and with gradients being all the rage nowadays for buttons and backgrounds, I figured I should probably use some of those to keep up with the cool kids. The first step was complete, I had my quick thumbnail.
But there were problems with the old blog besides the design, the greatest issue being that I seldom updated the it. For that matter, I wasn’t updating any of the three live domains on the net. Time was an issue, first due to school and then due to time spent looking after Little Nephew, but that wasn’t entirely it. I had to figure out why I wasn’t updating, and what I could do to address that problem. I decided to look at what I did and do participate in. I’m active a lot of social sites, from lj to myspace to facebook, but I also dump data into a lot of other conduits, like del.icio.us, last.fm, and twitter. While I have a decent number of friends on the pure social sites, I don’t really have many contacts on the others. Regardless of the networks, I had a lot of good data in those other sites that was kind of isolated in them. Facebook’s applications were a great way to gather my tweets, links, and music, but I still didn’t have control of the data. What I really wanted was some way to pull all this data into one location.
It was an issue of scattering. I had all these cool sites but no central location. I toyed with the idea over the summer of doing a tumblelog, but couldn’t decide on a platform. I was also fairly pleased with how del.icio.us worked, although I still wish there was a via form on the site to credit link referrers. As I toyed around with other options including pownce, I finally opened a twitter account, originally to follow Merlin Mann’s tweets, but then I became enamored of the interface. The 140 character limit became a challenge, like writing a great haiku or limerick. Between del.icio.us and twitter, I had two thirds of the tumblelog equation, just on separate sites. I looked through aggregators for WordPress but couldn’t really find something that quite fit with what I wanted to do. I gave up for the summer.
So as the end of 2007 was fast approaching, I looked for aggregators again and finally found one that worked fairly well for WordPress. As I was already subscribing to my del.icio.us links and my twitter tweets, I knew it would be trivial to then import them into my blog, rebranding them generically as links and blurbs. Last.fm proved more difficult, as the top artists for the week is only available in xml instead of rss or atom feeds. I toyed with yahoo pipes but couldn’t quite get the output in a format I liked. I’m still playing around with options, including writing a php or python script to handle the problem for me automatically.
From that point, I knew I could get all this stuff into WordPress, now it was an issue of figuring out how to set WordPress up to properly handle all this varied information. At this point, I decided that I wanted to include my flickr stream as well. At first I thought I was going to have to set up separate WordPress installations for each content stream. After some further research, I was able to dump each outside source into its own category. Armed with a few choice plugins and some php modifications, I was able to keep each category separate from the main post stream (or the blog, if you will).
My posting content designed and structured, it was time to do some css/xhtml/php editing. The only structural differences (that I can remember at this point) to the original template as far as design work was concerned consisted of moving the header and the insertion of the navbar/pages bar across the top of the page. Everything else was well-formatted. The css modifications were a bit trickier, and I ended up adding a few classes and id’s to get things how I wanted them.
At this point, I had automated content coming in from sources I update elsewhere and a design I was relatively pleased with, but there was still one issue. I didn’t really like having quicklinks/asides in the sidebar on every page as k2 did, and I really didn’t see the point in having my tweets on every page. A front page was needed, which could then point users to the blog or optionally other content. While the design on the social aggregator currently is close to my original concept, it is still in need of further refinement. However, it was functional and serving data to the index crawlers, the css could come later.
Now the pieces were together, but there was one issue left. The feeds only went back 20-30 posts, and I had over 300 links posted on del.icio.us. Thus began the process of manually entering all the old content so that I could have a complete reference to them. It took a while and nearly cost me my sanity, but I finished that up last night (on the evening of January 15th).
Now I have a functioning, hopefully aesthetically appealing site that I will want to update. There are still a few warts to deal with, but it’s only a matter of time before they’re cleaned up. The only real challenge left is keeping this thing filled with content.