Traction offers advanced Web logging

The Web is all about information sharing. As a result, the number of information-sharing technologies, such as content management systems and knowledge management systems, has dramatically expanded, providing agencies with many options to increase intelligence and productivity. One of the most recent and potentially complex additions to the panoply is the Web log.

Agency officials who specifically want to explore advanced Web-logging solutions that can securely serve the entire enterprise and business partners will want to add Traction Software Inc.'s Traction 3.1 enterprise Web logging solution to their short list. Already deployed at the Defense Department, Traction 3.1 uses Java and Extensible Markup Language technologies to easily enable authorized users to share information, such as e-mail, Web content and business documents. Traction 3.1 may be more costly than other Web-logging solutions, but its advanced labeling and searching capabilities are well suited to sites with dense information-sharing requirements.

We had no trouble installing Traction 3.1 on several Linux-based servers, and within minutes, we were accessing the browser-based administrative interface. After enabling Secure Sockets Layer encryption, we defined several projects within Traction 3.1. Projects are one method by which content is categorized in Traction 3.1, and they can be made visible to all users or specific ones. We created several test user accounts and gave them access to specific projects. Traction 3.1 supports Access Control Lists and integration with directory services, such as Lightweight Directory Access Protocol servers.

Traction 3.1's browser-based interface can also be customized via skins, a common feature in Web-based information-sharing solutions. A skin is an element of a graphical user interface that can be changed to alter the look of the interface without

affecting its functionality. The product comes with a few skins out of the box, but administrators can quickly create a custom look and feel for their sites. We created a skin that would be ideal for use on an agency's intranet. In this regard, the product matches other offerings, such as Miro International Pty Ltd.'s Mambo Server, very well.

After the initial setup, we logged in as test users employing a variety of client-side Web browsers. Traction 3.1 does a good job of supporting a diverse set of browsers, including Mozilla, Netscape Communications Corp. Navigator and Microsoft Corp. Internet Explorer, so staffers and business partners will not be locked into one. The solution also supports the AvantGo Inc. browser to enable access for mobile users.

Once logged in, the software's power becomes evident immediately. It was easy to add content using a Web form, and we published content to both private and public projects. Published content can be read, edited, classified and erased by

authorized users, all from the convenience of a Web browser.

Microsoft Windows users also can take advantage of another Traction tool called the Instant Publisher. Accessible via the Windows tray, Instant Publisher enables users to publish Web content from any URL, e-mail content from Microsoft Outlook and information from business documents such as Microsoft Word with a single click. However, Instant Publisher is tied only to Windows. Users of other platforms, such as Linux and Apple Computer Inc.'s Mac OS X, will have to utilize the Web form for publishing.

Agencies with dense information sets will find Traction 3.1's labeling and search capabilities first rate. The software comes with a default set of labels, such as News or Contact, which define the rules for content display. Any number of agency-specific labels also can be defined. We created a custom label called FAQ to mark content as available for frequently asked questions. Users can label entire articles or insert labels at the paragraph level for highly categorized content sets.

Labeling content may seem like a strange task at first glance, but once we began testing the software's search capabilities, the value of labels became apparent. For example, after we used the FAQ label for several articles and paragraphs, a quick search yielded all relevant content. Agencies executing research in one or more areas will find labeling and searching to be powerful tools in aggregating meaningful data.

Traction 3.1 also provides useful e-mail interfaces. For example, it was able to poll our test e-mail account and route incoming messages into a private Traction 3.1 project. Agency officials could use this feature to catalog and categorize incoming requests from the public or contractors.

We also used the software's outbound e-mail capability. E-mail digests of content activity within a Traction 3.1 site can be sent to authorized users at given intervals. We were able to verify that only specific users received the digests.

Traction 3.1 does not yet supply the broader pluggable framework found in leading content and knowledge management systems. Instead, it focuses specifically on the Web-logging aspect of content and knowledge management. Its powerful labeling and searching capabilities are ideal for sites that are swimming in content.

The array of information-sharing solutions available today offers administrators many choices. However, it also means that they need to carefully define the specific features required before attempting an evaluation or proof-of-concept project. For the greatest extensibility, a pluggable information-sharing framework should be considered. However, sites that need to tightly classify dense content sets via Web logging should definitely consider Traction 3.1.

Biggs, a senior engineer and freelance technical writer based in Northern California, is a regular Federal Computer Week analyst. She can be reached at maggiebiggs@acm.org.