Media Industry

CGI-ROM servers on a disc & other Web/CD intersections

Emedia Professional, March, 1998 by Ron Gustavson

While CD-ROM freezers its digital content in a polycarbonate snapshot at replication time, a connected CD-ROM--one which builds in an online or Web link--can evolve into a product that may little resemble the original title. The CD-ROM/online hybrid provides greater relevance to disc users for months after publication, and can provide a higher level of interaction, including transactions, database updates, integrated search and retrieval, and much more.

Most RecentMedia Articles

When it comes to connecting a CD-ROM to the Web, the simplest approach is to use a browser as the interface for the disc's HTML-formatted data, and to link the user to absolute URLs embedded on disc. Many CD-ROM/Web hybrid titles to a step further, and access Common Gateway Interface (CGI) programs on a Web server that can update database records, format reports in HTML for presentation in the user's browser, post messages to a threaded chat forum, count the number of visitors to the Web site, read the environment headers from the client's requests, and generally allows the Web to operate dynamically in real time.

The problem with a disc that makes a call to its Web server is that an active Web connection is required--a condition that is not always available or desirable. An emerging trend is to embed a proxy Web server on the disc itself, to be installed on the user's PC for rich CGI interactions without an online connection. A proxy Web server is like a middleman, functioning as an intermediary server which forwards requests from the client on to another server. In the context of a CD-ROM/online hybrid title, the local proxy server can appear to the client as the server until the actual Web server is needed, and a connection must be made. Embedded proxy servers will typically use the default local host IP address of 127.0.0.1 to handle such functions as local searches, using engines that were designed for the Web, or any dynamically-written HTML that might be created on-the-fly in response to a database query.

Emerging "portable" Web products--titles made with MarketScape's WebCD engine are probably the best known today--can now embed more than GIFs and JPEGs into the thinnest of thin servers: a CD-ROM. Whether installed on hard drive during a title's installation or operating as an ActiveX control or Java bean from the disc itself, live interaction between CD-ROM-based browser and proxy server and a Web server is now a general publishing option that is becoming realistic and relevant, as more potential users are becoming Internet-connected. And the ability of these disc-based component servers to create autonomous portable custom client/server environments may be an essential ingredient in the mass popularization of the Internet among the general public, especially if the convergence between TV and the Web is to become widespread.

CGI: AN INTERACTION INTRODUCTION

So what types of disc-based interaction might be possible using Common Gateway Interface Web communication? Though client-side-only processing is possible and rich, interaction with a Web-based server is still the final goal for most connected-CD transactions.

The HTTP 1.0 protocol supports three methods of CGI communication: GET, HEAD, and POST. GET is used to request information from the server in an encoded Uniform Resource Identifier (URI), which combines the server's URL with escape characters, directory paths, and parameters. This is the method most often used in search engine requests, where the search parameters are placed in the URI itself when sent to the server. The HEAD method requests only a response header from the server, which is useful for checking the validity of links and obtaining environment information. The POST method sends data to the server in the form of a header, plus a body of information which is usually in text or HTML format. This is the common method used to submit HTML Form input that often appears as a textbox on a Web page to allow data input to a database.

In the following HTML excerpts, the Cisco Connection CD-ROM accesses first the local search engine, and then the Web-based one--both using the HTTP GET method.

Perl, C, and UNIX shell scripts are traditionally used on the server to interpret the sent data and output either a data table for updating a database or HTML text for return to the client. Although still new, increasingly Visual Basic and Java are being used in server-side CGI processing. And things are due to change even more: the new proposed HTTP 1.1 protocol has also added the Trace, Put, Delete, and Upgrade methods to these three original requests.

Some early examples of CGI-ROM occur on searchable CD-ROM titles using a Web browser as the client. These include San Jose, California-based Cisco Systems, Inc.'s Cisco Connection CD-ROM, which employs Verity, Inc.'s topicSearch engine to search across either the CD-ROM or the Cisco Web site for the latest tech support and service relating to Cisco products. topicCD-Web Publisher installs a local host server, which runs on the user's machine to field these search queries. Encyclopedia Britannica's Britannica CD is another early CGI-ROM title that employs Verity's topicSearch engine to serve up encyclopedia pages on-the-fly from either the disc or the Web.