Blog Refresh

I have a semi-long standing tradition of doing a little refresh of my blog/website during the week before new years. Obviously I didn't make it this time, but I'm starting to think that it's about time for me to do something different anyway. And because I have to think through everything beforehand on the blog, I thought I'd collect a few thoughts here.

When I started tychoish.com, I was running a site at tealart.com [1] which was supposed to be a big umbrella site for a lot of projects, and I started tychoish as a place to collect regular thoughts and blog in a more "bloggy" sort of way. I felt at the time that TealArt (TA) had become too much of a magazine, and not light hearted enough.

When I gave up on TA, as it were, I was displeased with where the site was going and I felt like I needed a fresh start. I got that, and for all it's ups and downs, I'm really pleased with were tychoish has taken me. So this redesign isn't about needing to fix something that's broke, as much as it is to provide the opportunity to expand and explore additional options.

I am, admittedly a bit worried about expanding the focus of tychoish so that I open myself up to the "scope" problems that I had with TA. Nonetheless I'm not thinking about fundamentally changing what I do or what I write about, but more some look-and-feel modifications that would take this site from being "just another blog" to a "big tent" that would cover a lot of personal projects.

Because I'm a geek and my day job deals a lot with content management strategies, I'll start first with the way I envision the new landing or "front page" and then think about the technological solutions that I envision for this purpose.

Rather than have tychoish.com point to a blog, I think a sort of dynamic page that presents some content and links to other sub-sites (and more content), and has a layout that's different from other pages of the site is the way to go. I want to have the most recent "esssay" (my typical blog posts) an excerpt from the next essay, and then a list of links to the next 5 or so posts. I also want to have a few posts from the coda sub-blog, (longer forum microblogging and more "journal"-type posts).

I'm also thinking that it would be nice to include archive navigation, search, and other information that's now in the upper-right of the site, in the bottom, in a "hemmingway style" footer. I think that I could probably also link to and pull in content from other parts of the site, that represent projects that don't quite exist yet. I've been working on the foundations of a wiki-project that would be a resource collection for scholarship on open source stuff, and I think it would be good to pull in links to critical futures, and other projects. The blog would still live here, but other projects would as well.

Deeper in the site, the pages might look reasonably similar to the way that they do now, but those pages are mostly shorter anyway. One of my fears is that the entry page has too much content and not enough context, and I think by re-flowing the layout the site might make more sense. While the "wordpress-esque" look gives sites context, I'm not sure that it's the right context for tychoish. More deliberation and feedback on this is encouraged.

As for the backend, I'm toying with giving up Wordpress. While I've used wordpress for four years now (and I used b2/wordpress for a couple years about six or seven years ago; after using a custom CMS that Amy wrote for a couple of years), I'm starting to think that it's not quite right for what I'm trying to do.

I've realized that I need to write a post about monolithic and microkernel approaches to web-content management, I'm starting to think that tychoish isn't so much a site that needs a unified content management system, but rather a site that needs to be able to flexibly pull together content from a lot of different, but related, sources. I'm not talking about aggregation in the way that "tumblelogging" pulls together diverse connect, but more like I don't need one content management system that can generate my entire website. I'm pretty technologically adept after all, and I don't need a piece of software designed to "make my life easier" that doesn't quite conform to the way that I work.

So I mentioned last week that I was thinking about using a tool called "Jekyll" to power the essay section. I'm still following that train of thought, albeit slowly. Given that wordpress can import from RSS, I feel pretty safe that I won't be locked into anything. Truthfully the main thing I need to think about hacking in is possible support for multiple authors (easy, I just need to do some reading) and figuring out a way to draft posts (easy, just have a local branch, I think), and cue up posts for timed release (harder.) I've gotten to use post scheduling a fair bit in the last couple of months, and I could be convinced to not schedule posts except for times when I want there to be content and I'm not home.

Critical Futures will stay with Wordpress because while I think I can handle a much more ad-hoc website, if Critical Futures ever gets another editor I wouldn't want "massive UNIX chops" to be a required skill.

So here's a list of things that I need to do:

• Do some editing of the categories on my current wordpress directory
• Start using disqus comments with the current tychoish.com wordpress site, as that's what I'd be using in the new site.
• Explore Jekyll's wordpress importing feature, and potentially write my own wordpress theme to handle the export (with wget). Because I need some way to maintain the category data. Conversely, figure out how to get an RSS import into jekyll. Just have to see what's easies.
• Design a new landing page and hack around with CSS, because I'm my own design team.
• Cobble together something reliable that can aggregate RSS feeds from various parts of the site. Such a tool certainly exists, I just have to find it and get used to it.
• Possibly move to a different webhost. Long story.
• Implement the new landing page. I'm inclined to use PHP to pull together content (includes and what not), because that should be peppy enough without needing a database, and even though the backend would be ruby, PHP seems uniquely suited to handling the display-related functions. And I'm used to it. So there

Thoughts? Feedback? Cheerleading and/or other generous offers of support?

Onward and Upward!