The Harvard Weblog Project is providing anyone with a harvard.edu email address to create a weblog. One aspect that I haven't seen discussed is how people are going to find newly created weblogs.
At present, Dave Winer is maintaining a list of Harvard weblogs in Manila. I am guessing that this list is generated automatically because 1) the listing is alphabetic based on the first name, and 2) it will be a bear to maintain by hand as more people create weblogs.
When I created the De Anza Faculty Directory in Manila last year, I was constrained by the existing directory. This directory listed faculty in alphabetic order based on their last name and provided links to their sites. I needed to match this directory as closely as possible, allow it to continue linking to existing faculty sites and grow with new faculty Manila sites as they are added.
To achieve this, I added several fields to the basic user information that Manila stores. This included first and last name, a flag if they had and existing site and the url to the existing site. From these data, I generated a memberIndex table. Each entry in the index was named for the users lastname and first three letters of the first name. The entries either contain the canonical name of their Manila site or their email address. The faculty directory is created from this index. If an entry contains a canonical name, we create a link to the Manila website, if it contains an email address, we dive into the user information and grab the url from there.
As new sites are created, new users are added to the index.
A year later, with over 200 entries, the directory is a bit unwieldy. As I discussed in earlier posts, I have developed an alternate format that should allow easier navigation through the directory.
1:53:41 PM 
|
