First, a quick word - this essay should be considered at best a rough draft for now... I'm trying to put ideas together, and I'm dragging my feet.  But, I want to share these ideas with at least a few people to get some dialogue going.  A sister essay / wiki is located here.  Probably more polished versions of these ideas will end up there.

—

I've been talking with a number of people recently about the kinds of things that we'd like to see happen with content management systems.  As I see it there are two large segments that are currently underserved, and which could be treated with superficially different, but deeply similar techniques.  I've made some small progress on each of these issues, but there's plenty of road left to tread.

The first segment consists of people and organizations that do stuff and want others to know about it.  This could be classes, performances, meetings, private sessions, or anything else that occurs in a time and place.  One gold standard these days seems to be a service like constant contact.  These are nice, because they deal with all the complex mailing list issues (whitelisting, spam filters, etc.).  They also provide tracking of who's actually viewed the content you send out.  But then the only copy of the event is in e-mail form (which some people will lose), or otherwise you need to enter the same information twice to get it on a website.  Individuals may want things in different formats.  Social networking sites like Yahoo groups, MySpace and Tribe provide people with the option to see things as they come out, in a digest form or by visiting a website.  And everything is archived.  The sites, however, are structured quite rigidly and don't allow fine-grained control of who can post and change things.  They are certainly far from professional looking.  There are services like eVite, which send people an e-mail, then require them to go to a website in order to see what the e-mail is about.  Serious users of all of the above almost universally maintain separate websites which replicate all of the information that might be obtained on these more service-oriented web sites.  This provides a professional looking online presence, but requires additional replication of information. 

The second segment comprises people and organizationsre managing projects.  This domain is far more ranging and complex, probably because there is a great deal more money concerned about such things.  Heavy hitters in this category include things like SAP, and software you're expected to run yourself, like Trac.  But a popular newcomer is Basecamp.  All of these (well - I don't know very much about SAP, I must admit) include a central notion of projects and tasks which remain to accomplish these tasks.  There are mechanisms for actually storing the results of one's work, be it source code, documents or reports.  These tend to be restrictive and overly simple, or otherwise very complicated.  Sometimes they are a combination of both!

Central themes in both of the above are:

Lack of Control

Publishers lack a fine-grained mechanism for allowing individuals to post and edit information.  Readers usually have little control over the mechanism - like choice between e-mail or web-based delivery.

Loss of Connections

Different kinds of information generally go in different places.  There is often no mechanism for making arbitrary connections (not even using a hyperlink).  On-line systems are often designed to actively prevent viewing the information from outside the control of the service provider.

Starting at the opposite end, there are two basic formats that are fairly ubiquitous and easy to use on the web.  The first is the forum, stretching our notions, this would include e-mail as well as sites like blogs.  Here the organizing principle is time.  The second is the Wiki, which is a self-linking, generally flat system.  In this case the organizing structure is emergent.  Once we get offline, however, there are two additional structures which are worth noting.  One is the simple list, and the other (which perhaps contains the list) is the outline.

Final notions to mention are things like Google Desktop and Apple's Spotlight.  Search is not good enough, because it...

• only works if you know what you're looking for
• is subject to all of the issues of machine-based natural-language processing
• is useless to non-domain experts, or people who are new to the system
• fails to communicate connections

So, what do we need?  I like Zope very much, but more generally, and at a minimum, we just need something that is easy to use, and lets people keep structured lists around with things like time-stamps and links to other stuff in the system.  Surely, in our wild web 2.0 world, we can accomplish such a mean feat without too much engineering hassle.