Will the push -- not pull -- of Internet information dramatically alter our Web interactions? Learn why the next metamorphosis of the Web will be based on push technology and about the stimulating vendors who will lead us there

March  1997

Push technology is changing the way we access information. Not only does it alter the direction of information flow, it saves the ecology of the Internet by reducing traffic. We take a look at what makes it more useful than other access methods as well as the vendors who hold promise to make it so. (2,000 words)

Mail this article to a friend

The evolution of open information access technology has taken another step forward. With "information push" technologies coming from leading Internet software vendors, we are moving away from hunting towards gathering. As most know, the Web is an ever-changing system for information access. The way we access, view, and process information through the Internet has gone from static text documents to dynamic, graphical, and complex multiformat documents. Interestingly enough, the adoption of push technology also seems like a step backward from a different point of view. Let me tell you what I mean.

A very basic definition of push technology is the ability for the user to receive information passively from select sources over a network. Users install a push client package on their systems, usually in combination with their Web browser, and select the data sources they desire. The data source, typically a Web server, periodically transmits information to all the registered clients on the network. These data sources can provide a number of different categories of information -- often called "channels." When the client receives the information, it displays a synopsis in a window and waits for input from the user. From the various sources of information, the user can select the specific topic (or channel) of their choice. The full body of the information is then presented to them.

The name comes from the fact that instead of users actively seeking or pulling information from sources on the 'Net, they now passively receive select titles of data sources that are "pushed" from these servers. In most cases, the user still has to actively select and then download the full information from the server. Although in most cases the data accessed from these servers are pure information, some services also provide applications and programs that can be accessed this way.

The early success of push technology companies like PointCast Communications has led several other vendors to join the market -- most significantly Netscape and Microsoft. Additionally, the focus of related technologies, like intranet development, the network computing system (NCnet), and the transformation of groupware, has placed further value upon this technology.

Evolution of Internet information technology
Let's follow the brief but effervescent evolution of the information technology on the Internet. Originally, the File Transfer Protocol was the herd animal of the Internet carrying files of information across the global network. FTP, however, was independent of data formats and file structure other than ASCII and binary files. With Gopher you could access information which was organized and hyperlinked across the network, but it was still essentially text based. The Web brought sound and pictures to the data, and improvements on the Web soon resulted in video delivery, active images, dynamic data, and actual programs using Java and ActiveX.

The Web took one step back from Gopher, unfortunately, because it lacked any hierarchical organization. Information could be categorized by anyone in any format they chose. Although it allows greater freedom, it is also completely anarchic. With a small selection of information sources this isn't a problem, but as the size of the Web grew, this became unbearable.

So search engines emerged. They allowed users to come to a common location and search for locations (URLs) that contained information on specific topics of their choice. The search engines of course worked well but as the wealth of information grew, users typically started getting tens of thousands of sources from typing simple queries. Even though search engines have improved their search capabilities, this still resulted in an overwhelmingly large and discordant selection of data.

Then some companies grew on the idea that they could write configurable software to do the actual searching and sorting for users. This was the birth of intelligent agent technology. Users give these agents a criteria of what to look for on the Web. They may also be given suggestions on where to look on the Web, or how often to look. The faithful agent then spends the evening, while the user is away, to go hunting around the Internet for the information. Users return the next morning to find a selection of topics that their agents have located for them.

Now the Web is falling into a new category of actively distributed information. It is a step above agent technology -- not only is there an agent that looks for these sites, but there are also a select number of sites which provide comprehensive information sources.

For the Web site administrator this was a boon. When you put up new sites it becomes a chore having to register them with the search engines on the Internet; additionally, this registration process isn't perfect, and often your site comes out at the bottom of a list of 10,000 which can be quite disheartening. With push technology, rather than wait for users to come to Web sites, these sites are sending users titles of stories and data to attract and actively solicit users to their sites in an automated fashion.

Why push not pull?
There are many reasons why pushing information is better than actively pulling data. From the users' point of view, the information now comes readily to their screens without having to actively search for it -- not only that, but the quality of the information is probably higher; rather than having to trudge through thousands of personal user pages of questionable quality, the sources now come from well-known information sources like the NYSE or The Wall Street Journal. Additionally, the data is already cross-referenced with other pieces of data on relating stories and news. It's essentially like watching a TV channel with the exception that the type of shows you see are already pre-selected automatically dependent on the criteria that you originally gave.

For the Web site administrator or marketer this becomes a jewel. They no longer have to beg users to come to their sites. Instead, they just start a channel with a specific service, and the information is delivered to a large audience without too much active promotion.

This is also good from the Internet's point of view in that now information can be presented in a categorized fashion. Users no longer have to go around the Web downloading pages just to see if the information is of any use to them. Less time is spent hunting around by the user, so the volume of traffic is reduced partially. It makes the user feel better, the Web site administrator more comfortable, and the backbones and connections more relaxed.

Broadcast media all over again?
The immediate result of push technology is that the Web now shares more in common with traditional broadcasting technology. TV, radio, and even the cinema, have always been broadcast because of the cost of the technology and limitations of bandwidth available across the world.

Broadcast technology has also led to the focus of information production to a select few organizations. In the US, most of the TV audience has access to news and information from a small group of sources -- primarily the major TV networks: NBC, ABC, Fox, and CBS. The merits of the TV broadcasting industry have been in question for quite a while but very little has changed over the past 50 years.

When the Internet blossomed, the ability to directly access the kind of information you prefer was one of the core merits; not only that, but the information was usually available to you when you wanted to access it. This is different from regularly scheduled programming on radio and TV. The Web is more of a narrow-casting scenario when users are fed their own selections of information compared to mass distribution of the same program to all users without independent access control.

Even though it is narrow-casting, you can't shake the notion that the same data on many Web sites is accessed by a large audience of people. The only difference is the fact that access to the Web information (thank goodness) is not all at the same time. It's true that every user will not be accessing identical sites at different times; some users will go to sites that others wouldn't be caught dead in for that matter. However, when you average it out, more times than not, you do get specific audiences at various sites.

With push technology, however, you change the face of information delivery on the 'Net. Now you do have a specific user audience (all those using your client software), and you have select programming of information. The technology only matches up one group with the other in a semi-efficient manner.

The next major network broadcasters
The vendors developing the software for push technology are becoming the equivalent of network broadcasting companies. These vendors include PointCast, Wayfarer Communications, BackWeb Technologies, Intermind, and Data Channel Corp. Additionally Marimba, the darling of the Java industry, stands close with its Castanet application push technology. Looming over the industry, giants like Netscape and Microsoft are also targeting this sector with products of their own.

PointCast was one of the earliest with its stock information and corporate news distribution system through special client software which can also act as a screensaver. PointCast has custom software for PCs and Macintosh systems that access its own servers. Collecting information from various sources, it provides a directional broadcasting system to a wide audience, primarily corporate.

Although Kim Polese's Marimba started with different product idea, the company now provides a technology which gives skin and skeleton to the promise of networked Java application delivery. While the concept that Java can provide portable code is a nice idea, the Castanet technology actually makes it possible to create applications which auto-install and auto-update from select "application transmitters" on the 'Net. With such a distribution system, users no longer have to even think about what they must do to prepare their systems to run a new application. It takes away all the worries PC users have had in the past.

I saw BackWeb at the Internet World Expo last December, and it proved to be somewhat interesting. McAfee Associates, one of the premiere makers of virus protection software, has teamed with BackWeb to deliver updates to its VirusScan 95 product through this new push vendor.

Almost all these vendors have released products for intranets to create corporate-local push services. Prices for these packages average between $25 to $200 per seat. Although limited by hardware capabilities, the server software can usually support several hundred simultaneous users.

While many are waiting to see what Microsoft and Netscape will do, these smaller companies are establishing themselves as leaders of push technology. Microsoft plans to integrate push technology into its Internet Explorer browser; essentially, it will integrate the PointCast client into Internet Explorer 4.0 towards the second half of the year. Netscape has been planning its Constellation "network desktop" for quite a while now. Although it provides several other features, it will also provide pushes of data directly to desktops as floating clips of news. Constellation is targeted for release near the end of the year although Netscape plans to release beta versions this summer.

We are watching the Web transform into a different animal. Information, rather than being hunted like exotic prize game animals, is now being delivered to you as quickly and simply as duck soup from your local Chinese restaurant. The vendors of push technology are joining forces with content providers to deliver on the promise of quality information to clients. At the moment they are still lacking sufficient programming, but you can be sure that this is a crucial time to establish themselves as leaders for the next generation of Web users.

Resources

• Marimba
  http://www.marimba.com/
• PointCast
  http://www.pointcast.com/
• BackWeb
  http://www.backweb.com/
• Wayfarer
  http://www.wayfarer.com/
• Intermind
  http://www.intermind.com/
• DataChannel
  http://www.datachannel.com/
• Netscape Constellation Preview
  http://home.netscape.com/comprod/tech_preview/index.html
• Microsoft's announcement on the integration of PointCast client into Internet Explorer 4.0
  http://www.microsoft.com/corpinfo/press/1996/dec96/ptcastpr.htm
• More on PC connectivity
  /sunworldonline/common/swol-siteindex.html#pcconnectivity
• A list of other Connectivity
  /sunworldonline/common/swol-backissues-columns.html#connectivity

About the author
Rawn Shah is vice president of RTD Systems & Networking Inc., a Tucson, AZ-based network consultancy and integrator. Reach Rawn at rawn.shah@sunworld.com.

If you have technical problems with this magazine, contact webmaster@sunworld.com

URL: http://www.sunworld.com/swol-03-1997/swol-03-connectivity.html
Last modified: