GROUPWARE
Prediction: The line between proprietary groupware products and standards-based intranet applications is rapidly fading and may well completely disappear from the user's point of view.
As we enter 1997, identifying exactly what is and isn't groupware is growing more and more difficult. Lotus shipped Notes 4.5 in mid -December, and Microsoft's Exchange 5.0 and Novell's GroupWise 5.1 will ship early this year.
All these products provide essentially the same functionality--calendaring, E-mail, document sharing, and other applications--to Web-browser users as well as to proprietary clients. Novell extended GroupWise's functionality to browsers in version 5.0; in version 5.1, Novell will include a universal client--a full-blown GroupWise client, with the user interface and all functionality written in Java--that can be used by browsers and network computers. Lotus also will introduce Weblicator, a client-side proxy that works with a Web browser and gives users offline access to cached Web pages.
Vendors of intranet and groupware-like applications such as Open Mind from Attachmate in Bellevue, Wash., and Intranet Genie from Frontier Technologies in Mequon, Wis., may find this to be both good news and bad news. On the one hand, the wholesale acceptance of intranet technology by the groupware giants validates their approach. On the other hand, those giants might absorb the specialized vendors' markets as well. The survival of specialized vendors will probably depend on three strategies: beating the big guys with more features, easier deployment, and better application-development tools; offering enhancements to the products from the groupware behemoths; or outdoing them on price. The first two hold some promise, but the third is close to impossible against Microsoft.
Prediction: Lotus Notes will continue to dominate the groupware market.
Lotus Notes leads in groupware component technology,development tools, and an army of third-party supporters. Although Microsoft has tried hard to position Exchange as anything other than a competitor to Notes, Microsoft attempts to match Notes feature-for-feature and packs every Internet standard into Exchange 5.0.
But both Exchange and GroupWise still lack the one technology at Notes' core--a database. Although GroupWise will add hard-core imaging features this year, and Microsoft has improved the integration of Exchange with the rest of its Back Office suite, Notes still maintains the critical advantage: a single package for groupware application development.
Meanwhile, mainframe and minicomputer mavens will probably see IBM make further moves toward bringing Notes to MVS and the RS/6000 SP platform, as the final successor to its venerable Profs office automation software. Integration of Notes applications with legacy and client-server applications could give IBM an easy way to merge those applications into the intranet while keeping security and high availability intact.
There remains one unknown quantity: Java. How well each vendor takes advantage of the power of Java could change the complexion of group computing.
Sean Gallagher is a senior technology editor for InformationWeek. He can be reached at sgallagh@cmp.com .
INTERNET DESKTOP
Prediction: Us ers will balk at fatter desktop applications.
The beginning of 1997 is looking a lot like the beginning of 1996. While new desktop software and Internet infrastructure technologies are making the Internet a richer experience for users, bandwidth is as scarce as ever.
Where bandwidth is sometimes easier to find--in corporate intranets--desktop resources can barely keep up with fatter applications. Unless new versions of software require less bandwidth or offer significant productivity gains, fewer users will pay for an upgrade.
Enhanced versions of the most popular browsers are, of course, inevitable. Each revision of Microsoft's Internet Explorer and Netscape's Navigator brings richer HTML implementations, better Java support, and the usual extra whiz-bang features. Unfortunately, they also bring increased resource requirements.
With the current 3.0 browsers already requiring a 16-Mbyte machine and a Pentium or PowerPC CPU for good performance, you can bet that 1997 wil l be the year that many corporate desktops don't upgrade to version 4.0.
Prediction: Multimedia over the Internet will still be reserved for the bandwidth-privileged.
Audio and video delivery matured in 1996, but bandwidth and desktop resource limitations continue to make it problematic in 1997. Products such as RealAudio from Progressive Network in Seattle and VivoActive from Vivo Software in Waltham, Mass., significantly squeeze streams of audio and video data, but it's often not enough. In general, only the privileged few with ISDN (128 Kbps) or faster connections enjoy consistently clean sound and video displays. This is not likely to change in 1997.
Meanwhile, Microsoft's NetMeeting and the forthcoming Netscape Conference conferencing software make multiperson, text-only collaboration easy for even low-end users with 14.4-Kbps modems. And for high-end users, rich "virtual community" software, such as CyberHub from Black Sun Interactive in San Francisco, can bring sophistica ted visual collaboration into corporate domains using VRML. But it probably will be a bit too resource-heavy for many sites.
Prediction: Products that filter and distill the chaos of the Internet into usable chunks will be big.
One resource-heavy application area that many users are willing to tolerate are so-called "push" technologies that broadcast the latest news or Web-site updates to desktops. Products such as PointCast from PointCast Inc. in Cupertino, Calif., and Intermind Communicator from Intermind Corp. in Seattle let users pick the sources and types of information they want to see.
Jason Levitt is a senior technology editor for InformationWeek. He can be reached at jlevitt@cmp.com .
WEB DEVELOPMENT
Prediction: Developers of Web applications will be forced to align with either Microsoft or Netscape.
The lines between the factions creating Web application tools and infrastructure are becoming more defined and harder to straddle. Microsoft and Netscape both provide the client and server platforms on which the majority of Web applications run. Through this year, Microsoft will be enhancing and improving its Visual J++ for building Java applets and its Front Page 97 Web authoring tool for site development. The upcoming release of Visual InterDev (formerly known as Internet Studio) will fill out a set of Web development tools that are unmatched for power and consistency.
But there is a catch. That power and consistency are largely due to strong platform ties. More than before, developers will have to choose between the elegance and high performance of platform-dependent products and the versatility of cross-platform solutions.
In 1997, Microsoft will pull together a well-synchronized Web development package: Windows NT, Internet Information Server Web server, ActiveX component objects, Active Server Pages that run JavaScript or VBScript on the server, Visual InterDev, Front Page 9 7, Visual J++, Visual Basic 4.0, and more. These tools will shine together on Microsoft Windows platforms, but they will not reach full potential separately in cross-platform environments.
Especially problematic will be Microsoft's level of adherence to the Java language, class library, and Virtual Machine standards set by JavaSoft. Microsoft's version of the Java VM for 32-bit Windows already differs significantly from the standard (more by the addition of ActiveX than by omission of any feature). Microsoft has not yet committed to support all of JDK 1.1 and will likely support the JavaBeans component standard only in areas that are consistent with its ActiveX and DCOM (Distributed Component Object Model).
Another faction is forming around Netscape and its key development initiative, Netscape One. Netscape is using its Web development architecture, based on Java, JavaScript, and the Web protocols (HTTP, SMTP, etc.), to practically declare war on ActiveX and DCOM. This year, more tool vend ors and users will pledge allegiance to Netscape's point of view, joining Symantec in Cupertino, Calif., with its Visual Café Pro development system; NetDynamics Inc. in Menlo Park, Calif., with its NetDynamics Web database access tool; and others.
Unfortunately, Netscape One is also veering from the JavaSoft standards. To overcome the constraints of Java's limited user-interface class libraries, Netscape and its partners will push its Windowing Internet Foundation Classes (IFC) as well as introduce other IFC libraries. These libraries are Java classes beyond the JavaSoft standard. IFC will be included in the next generation of Netscape's browsers but will have to be installed into existing browsers. Time will tell how well IFC integrates with JavaSoft's new class libraries. We'll also have to see how Netscape's Java Virtual Machine wavers from the standard.
Some vendors of Web development tools will align with one faction or the other; others will ship one version for Netscape On e and another that works with Microsoft's ActiveX and DCOM.
Jeff Jurvis is a senior Internet consultant at Solid Logic Computer Solutions Inc. in Eden Prairie, Minn. He can be reached at jeff@slogic.com .
MANAGING NETWORKS
Prediction: The Internet craze will add two main responsibilities to the network manager's job: monitoring traffic into the corporate network, and creating an intranet.
Constant monitoring of outside information coming through your company's Internet connection is critical. The two primary areas of concern are outsiders getting inside, and insiders bringing in viruses or inappropriate material.
The hacker threat has been in the news so often that no company should dream of adding an Internet connection without some type of firewall installed. Security is never a "plug and forget" operation, however. Monitoring the firewall and other security systems is a daily priority, because just a few hou rs of downtime on your security system can spell disaster.
True, it's unlikely you'll be attacked during the two hours on a Tuesday afternoon that your firewall crashed, but any exposure will only aggravate your ulcers. You'll also need to constantly update the firewall to reflect the changing methods of random hackers and corporate spies.
File transfers into your company, even when initiated by trusted employees, create more network problems. Viruses within companies are on the rise as employees succumb to the lure of new applications on the Internet. Downloaded files from major shareware and corporate Web sites are usually clean, but "Bubba's Freeware Server" may not be. Add virus filtering to your firewall if possible, or restrict the number of users who are allowed to download files.
Inappropriate material (primarily pornography) will land the company in hot water if downloaded and spread around. Blocking access to certain newsgroups (alt.*) and Web sites will become mandator y, as will random checks of downloaded material to verify that only business-related files are introduced.
The network manager's second major problem is duplicating the Internet environment on the internal network in order to turn a LAN into an intranet.
Again, there are two prongs to this project: your choice of client interface software and internal network protocol. Users are being pushed to accept the browser as the universal client interface, and that does make some sense. Internal applications must be updated to reflect this thinking, and new projects should be Web-oriented from the beginning.
Although Novell's IPX protocol connects two-thirds of corporate desktops today, intranets demand TCP/IP. You will be forced to change each NetWare client to support TCP/IP directly or add gateways. Novell's IntranetWare (NetWare 4.11) includes an IPX-to-TCP/IP gateway, as well as a solid Web server for intranet use. At least a dozen companies offer competing gateways, most of them wi th more features than Novell's.
Of course, your carefully crafted security profiles on the internal network will be worthless as you convert to an intranet. The bad news is that internal firewalls may become just as important as external firewalls. The good news is that these changes guarantee experienced network managers continued employment, and probably larger than average raises, throughout 1997.
James E. Gaskin is a network consultant, columnist, and author in Dallas. He can be reached at james@gaskin.com .
DATA WAREHOUSING
Prediction: Data marts will gain momentum in the market as commodities.
In 1996, the breathless enthusiasm for data warehousing was tempered by troubling stories of projects foundering or lost, evidenced by the occasional flotsam of missing deliverables, disappearing benefits, and damaged careers.
The industry responded with a "new and improved" approach: the data mart, which is reputed to be more manageable because of its reduced scale. Expect this trend to continue and grow, although data marts are only a part of the data warehousing phenomenon, not a replacement for it. Corporatewide data warehouses will hold the high ground for serious decision-support applications, for analysis of cross-functional enterprise views, and the beginnings of the "public" data warehouses as a revenue generator.
Prediction: The consultant charlatans will ride off into the sunset. Only quality products and services will be in demand.
Data warehousing has a long past but a very short history. It's surprising, then, that there is so much so-called experience out there. But times of great change always see a proliferation of shamans and medicine men, a major factor in the often bleak returns on investment for a large number of projects.
Buyers of this technology will no longer tolerate the endless parade of heroic gunslingers and their saddle-sore solutions. The "early maj ority" buyers insist on referenceable success in like situations, realistic and quantifiable value propositions, and a track record of promises kept. Buyers of products and services will be more discriminating, and those in the data-warehousing business not making the cut will cycle back to the comfort of early-adopter markets like data mining, visualization, and the Web.
Prediction: Microsoft will emerge as a serious data-warehousing contender, with offerings in databases, operating systems, and online analytical processing tools.
The maturing data-warehouse market will also exert a decidedly negative influence by creating de facto standards based on vendor name recognition. Those vendors with superior products who haven't become household words will slip into obscurity. The bulk of the money spent will be concentrated in a few name-brand hardware, software, and systems-integration vendors.
The situation doesn't preclude smaller companies, such as Arbor Software, Business Obj ects, Cognos, MicroStrategy, and Red Brick Systems, from gaining disproportionate amounts of mindshare and achieving accelerated growth. But their share of the market will likely remain small.
Prediction: No data warehouse will be complete without a Web interface.
Like everything else in the technology business, data warehousing will see more of the Web. This will be integrated into the reworking of early warehouse projects. Many were designed five or more years ago, predating many of today's exciting technologies. This augurs well for smaller, better vendors that will finally have an informed audience to hear their story.
Neil Raden is president of Archer Decision Sciences, a data warehousing consulting company in Santa Barbara, Calif. He can be reached at nraden@ix.netcom.com .
OLAP TOOLS
Prediction: The Internet browser will be the de facto user interface for analysis.
As demand has increased for query and report writers, report servers, and online analytical processing (OLAP) tools, so too has the competition. Many vendors are branching out with sophisticated products packed with analysis capabilities, report servers, and data-mining tools.
But as with everything else, these tools are headed straight for the Internet.
The Internet and intranets make wonderful mechanisms for distributing reports, both internally and externally. A number of major query and reporting tools already support HTML as an output format. This was relatively easy to accomplish; the more difficult task--and what few have done--is letting the user run the query tool from a browser.
The primary way to add a Web-browser interface is to turn the product into a Netscape plug-in or Java applet, which is dynamically downloaded from the vendor's Web page. This architectural shift will have a profound impact on the cost of query and reporting products. With the vendor charging according to the number of servers, the incremental cost of adding a user will be close to nothing.
Also affected will be the way query and reporting systems are managed. Mainly, there will be a greater emphasis on central administration. This will let companies easily distribute reports to a new class of users--their customers.
Jay Tyo is an analyst at Rochester Gas and Electric in Rochester, N.Y. He can be reached at tyo@rge.com .
DATABASES
Prediction: Multimedia data will steadily creep into operational databases.
1997 is the year users can see any type of data any way they want.
New applications will be enabled by the burgeoning object relational databases that store multimedia data alongside names, addresses, and Social Security numbers; Web database interfaces that deliver data dynamically in HTML pages; and large-capacity disks that can hold the corpulent multimedia files. Corporate developers will be able to give users the rich interface they demand: clear document images retrieved in seconds, animated Web pages with colorful images, and annotated maps and diagrams.
Julie Anderson is technology director for InformationWeek. She can be reached at janderso@cmp.com .
APPLICATION SUITES
Prediction: Office productivity suites will be dominated by four words: collaboration, Internet, easier, and bigger.
Microsoft, the dominant force with more than 85% of the application-suite market, has already released its Office 97; Lotus will soon counter with SmartSuite 97, and Corel will release Office version 8 this spring, followed by a Java implementation this summer.
Previous releases of Microsoft Office, like the other products in this category, had limited team support. SmartSuite 96 changed that, with excellent team features, including the ability to easily route documents to other users for comments and changes. Each user can make change s to a document and route it back to its originator, who in turn can accept or reject the changes.
SmartSuite 97 will extend this collaboration to the 1-2-3 spreadsheet. Corel's Office has not focused on collaboration in the past but will have more document-sharing features in version 8.
Integration with the Internet is the chief goal for suites in 1997. All the suites will let users publish documents and presentations to the Internet as well as embed hypertext links in documents or presentations.
Corel is also taking the plunge by creating an implementation of Corel Office using Java. But the immaturity--as well as system memory and CPU demands--of the current Java run-time environment will limit the functionality of this implementation. For simple tasks, such as creating interoffice memos or small spreadsheets, Office for Java will be a good fit for those companies moving toward the network computing model.
All of the suites will have wizards, interactive help, intell igent assistance, and natural language to make it easier to create documents and presentations. But all these cool new features add size to the code, which makes the suites bigger.
Microsoft's research suggests that folks are using only about 20% of the functionality in office productivity suites. But users are also demanding new features. This seeming contradiction will result in ever-expanding disk requirements.
Vendors should break their applications into finer-grained components and allow users to decide which pieces to load. For example, why not turn version control into a component that's loaded only if the user wants it? Look for Corel to start the ball rolling on fine-grain feature installation in its Office version 8.
Andy Feibus is president of AMF Enterprises, an Atlanta consulting firm specializing in application development. He can be reached at amf@mindspring.com .
http://www.informationweek.com