Biografia e approfondimenti
Internet Based - Collaboration in 2010
Imagine a common scene: A small group of people from
different organizations meets and decides to engage in a cooperative project.
These might be people working on a business deal, people sharing technical
information, or a social group planning a future function. They decide
to use the Internet to facilitate their interaction. What technology are
they likely to use? E-mail is certainly available, but there is currently
little else. The only other widely used collaboration technologies are
AIM (AOL's Instant Messenger service) and ICQ, which are quite useful
but primarily for short interactions. They provide no tools for recording
the interaction or connecting the communication with the users' other
files.
After 30 years of vigorous technical developments and an explosive adoption
of technology, this strikes me as embarrassing. Fortunately, coming developments
will trigger dramatic changes in collaboration technology and systems.
The Technical Drivers
I write as 1999 ends, and I see five technical developments that are laying
the foundation for a radically different future for Internet-based collaboration.
Security. Control of access to the information is vital to collaboration
over the Internet. Throughout the Internet's history, government controls
and patent restrictions have limited the adoption of high-quality, fine-grained,
security mechanisms. The patent on the most important of the cryptographic
algorithms, RSA, expires in September 2000 and will open the door to broader
scale adoption of this algorithm in all levels of systems. At the same
time, the U.S. government's long-standing ban on exporting systems with
strong encryption appears to be weakening. Once these encryption and authentication
technologies are widely and easily available, new architectures will emerge
that incorporate fine-grained security control at every level.
Bandwidth. The vast majority of users today are connected to the Internet via dial-up lines. Large- and medium-sized enterprises fare better with T-1 or even T-3 connections, but they account for only a small fraction of total users. New "last mile" technologies, particularly DSL and cable modem, are rolling out and will become widespread over the next decade. These technologies bring higher bandwidth at a modest price, and they also allow the user to be always connected to the Internet. For collaboration, this will facilitate services that provide a sense of proximity and immediacy.
Convergence of voice and IP. Today the telephone system--often called "POTS" or plain old telephone service--and the Internet are disjoint, essentially unrelated systems. However, the major telephone companies and major equipment vendors (such as Cisco) are actively developing merged services running voice over IP (VoIP). There is substantial work left to assure quality of service, but one sure effect of VoIP will be to integrate the movement of data and voice. This will provide a natural foundation for building collaborative services, such as for sharing documents during a teleconference session.
Appliances and handheld devices. Personal digital assistants are becoming commonplace. The general-purpose desktop computer will lose ground to more specialized, and easier to use, network "appliances." Over the next few years we're likely to see a variety of designs, many of which won't survive, but by the end of the decade these appliances will become the dominant means of accessing the Internet. People will become dependent on continuous connection to the Internet as an integral part of the computing systems they use. As part of this, they'll also rely on interaction with others and seamless document sharing.
Wireless connections. Today, there are relatively few mobile Internet users: Most mobile service is layered on top of cellular telephone service and is relatively expensive and slow. Over the next decade, wireless connections will become more common and much less expensive. Metricom's Ricochet service currently operates in a few cities at dial-up speeds, and it--or competitors--will evolve to offer much higher speeds in many, if not most, metropolitan areas. Along with the trend toward building easy-to-use network appliances, we will see a large number of mobile users in continuous connection with the Internet.
Let's Get Together
Group interactions will take many forms. Within the next decade, I expect
to see widespread sharing of documents, sharing of real-time drawings
on whiteboards, voice conferencing, and, perhaps, videoconferencing.
Document sharing. Today, people share documents primarily by sending them
as e-mail attachments. Within large enterprises, systems like Lotus Notes
provide a means for sharing documents through common directories and replications.
At present, however, these systems are too expensive for small businesses
and individual users; the architecture is inadequate across organizational
boundaries. In the future, users will share documents as easily as they
now send mail or make phone calls. They will simply specify whether the
sharing will be of a snapshot of a document or update versions as changes
occur.
Diagrams and doodles. Sharing of drawings will explode. Today, there is rudimentary use of NetMeeting and similar tools for sharing drawings in real time across the Internet. Those tools are not robust, do not scale well, and are not integrated into the rest of the Internet protocols, but there are few technological barriers to improving them. In settings as formal as a boardroom and as casual as a coffee shop, it will be possible to immediately capture drawings and share them with colleagues across the world.
Voice conferencing. Phone calls are commonplace, but conference calls are difficult and expensive to arrange. The POTS system is not engineered or managed to make group conversations easy. The transition to IP transport will facilitate a change and make it possible to share conversations among multiple parties more easily and at lower cost.
Videoconferencing. The videophone has been dreamed-of
for decades but hasn't yet emerged. However, the use of strong compression
algorithms, the sharp reduction in the cost of cameras and computing bandwidth,
and the increase in communication bandwidth all suggest videoconferencing
may transition from an expensive, point-to-point, by-appointment-only
service to a ubiquitous, on-demand service available to everyone.
Prognosticating about the future is hazardous. A decade from now I anticipate
rereading these predictions with regret at how far from the mark I was.
But I expect to be able to share the experience over the Internet with
my family and colleagues wherever they happen to be.