Getting started

I've decided to start a more general-purpose weblog.

I’ve decided, with some encouragement last summer from Eve Maler, to start a general-purpose weblog. I’ve done a lot of writing on technical topics in various media over the years, but postponed doing this until I felt more of a sense of purpose about it.

My use of blogs so far has been more experimental:

  • Thinking About Linking on the O’Reilly Network focused on linking-related issues. I began it in April of 2003 when a now-defunct mailing list on hypertext fizzled, thinking that I would write my ideas about linking and then whoever wanted to read them could when I wrote them or at some point in the future. My plan was for the list of postings to become a resource for myself and other people doing research on the topic, and that hopefully I’d get some feedback on those ideas. I also wanted to see if greater focus in a weblog made it more valuable than one about just anything.

  • sneetard is a blog by and for two people, so that my brother and I can point out funny things we’ve seen to each other. This topic formed the bulk of our e-mail before (much to my wife’s frustration when I told her that Peter and I e-mailed each other three or four times on a given day and she then asked “how is he?” and I had no reply), and the blog reduces the chance that one of us will point out something on WFMU’s Beware the Blog or boingboing that the other has already seen.

  • Xanadu is a private weblog for about 45 people involved in metadata and the publishing industry.

What all three of these have in common is that none is a blog in which one person writes about anything they want for the whole world to see, like most weblogs. Each has a fairly specific purpose, goofy or otherwise. will be a lot closer to the one-person-soapbox model, but I will continue to refrain from discussing my new favorite CD or what I had for breakfast. I’ll be discussing new and old technology, and a weblog on my own domain, in which I can tweak the stylesheets and play with the settings to understand that technology better, is a more appropriate place to do that. As to old technology, I’ll paste and edit the last two paragraphs from my last O’Reilly weblog posting here:

I’ve learned from recent reading that most histories of computers focus on computers that specialized in the most advanced math possible for their time, which was as much of a niche application in 1900 and 1945 as it is now. Many key tasks that we use computers for today—particularly database tasks—were being carried out by automated, usually electrical machines since the nineteenth century in a separate but parallel history to the Collosus-Mark I-ENIAC-EDVAC history of computers that you typically read about. Did you know that during World War I the U.S. Army could run automated queries against a database to find, for example, French-speaking soldiers with a chauffeur’s license? Lately, I’ve been fascinated by large-scale database applications that predate any database technology that geeks currently take seriously. A lot of people now consider any pre-relational technology to be prehistoric; that’s a pretty limited perspective.

The history of computing applications, with or without the use of electrical stored-program computers, has a lot to teach us about the problems and innovations we’re working on now. I’m sure I’ll be spouting opinions on other developments as well, especially XML-related ones, which I’ve worked with and written about since the days when XML was a four-letter word. (For example: now that the W3C “binary XML” effort has been renamed “efficient XML interchange,” there’s no reason to argue with it anymore, because who can argue with greater efficiency? Right?) When I see interesting linking-related news, I’ll probably add new entries to the O’Reilly weblog, by my main weblog from now on will be this one. I hope it’s worth reading.


By Aristotle Pagaltzis on December 4, 2005 6:02 PM

Hi Bob,

two requests on technicalia:

  1. Since it’s now / rather than /, could you please update your templates so the link tags in your page headers will point to the correct location? I picked “Atom” from there and got subscribed to the wrong feed, then had to manually fix it.

  2. The right way to handle existing subscribers, assuming you have enough control on the server to instutite it, would be to send a “301 Moved Permanently” redirect in response to requests for /, rather than putting a dummy feed there. Aggregators can then automatically update their subscription to the correct location.

Anyway, I look forward to reading you here as much as I did on O’Reilly’s site.

By Bob on December 5, 2005 9:55 AM

Sorry, I’m still getting to know what Movable Type puts where in the various templates. I just fixed it in the templates that generate the Atom file, index.html, and the various archive html files, so please let me know if you find any references to atom.xml elsewhere.

The 301 return code is a good idea, but I don’t have that level of control over the HTTP sent by my host provider’s server. I was happy to learn about .htaccess, which is how I get Atom files sent with a MIME type of atom+xml. In fact, that’s why I renamed the atom feed file to have an extension of atom.



By Aristotle Pagaltzis on December 5, 2005 10:08 PM

No problem, just pointing things out. :-) If you can create a .htaccess, you probably *can* send a 301:

RedirectMatch 301 ^/\.xml$ /

That would go in the .htaccess in your document root.

By Martin on December 6, 2005 5:22 AM

Do you have any plans on providing a full-content feed? I’d appreciate.

By Bob DuCharme on December 6, 2005 7:49 AM

As a matter of fact I was planning on doing that after I get the next post up, which is a pretty long one. Can you point me to any examples of sites that provide both a summary Atom feed and a full one? I want to make sure I get all the link elements in the secondary one right.

By Aristotle Pagaltzis on December 7, 2005 9:08 AM

Bob: I assume you mean autodiscovery links? In that case, you’d do something like


By Bob on December 8, 2005 12:54 PM

No, I meant actual websites. is about the only one I’ve seen, but I didn’t look very hard.

I now have an Atom feed with full entries at and will be publcizing it more in a blog entry.