There are different approaches to sharing documents, where a group of people want to be able to make comments on a document and alter it.

One approach is Track-Changes, which we’re all familiar with in MS Word. This is very useful. But then, after three or four sets of changes, it gets too confusing. The Track-Changes approach is best between two people; it does a reasonable job of letting you see what changes have been made.

Another approach is one we’ve just been looking at, that of SharedDoc, from SharedBook. Here, someone owns/controls a document. Other people can post responses at arbitrary points to comment on the content. Subresponses of arbitrary depth are allowed. Here, comments aren’t applied, unless the owner specifically decides to allow them. Reviewers must be specifically selected by the owner. This is another approach valuable in a variety of contexts where one seeks reviewer input, such as proposals, legal services, business documents, and publishing. SharedDoc is in beta, by the way, and is currently free.

Some other approaches to editing shared documents include:

• Wikis, where people can go in and make changes directly, with an audit trail of the changes. You find this approach in Google Wave.
• Content management-style versioning. You also find this in SharePoint, for example.

… David Ferris, with thanks to SharedBooks’ Caroline Vanderlip for her interesting insights

Conferencing technology–such as Cisco WebEx and Microsoft Live Meeting–has two approaches to the audio element. You can route it over the data network, or you can route it over the conventional telephone system (“PSTN” or “Public Switched Telephone Network”).

In principle, it’s much better to treat the audio as just another data type for the conference and route it over a data network. This makes things much easier for users, because you can have tight integration with the conferencing application. No remembering and entering phone numbers, for example. And usually cheaper.

Internally, companies often run the audio part of conferencing over their in-house data network, because their internal network is fast enough and delivers data in a timely way. However, where participants must connect over public data networks, notably the public Internet, it’s common to fall back on conventional telephony, because voice transmission over public data network connections often is not quite good enough for the audio part of a conferencing session (even though most of us have experienced pretty good voice quality at times via Skype).

There are two reasons for this audio-over-Internet shortfall:

• Voice over IP technologies like Skype convey voices much less well when several people are chiming in and even talking over one another–as quite often happens in a conferencing session — than they do for a one-to-one conversation
• Conferencing behavior is very sensitive to slight delays in delivering the voice signal voice, known as “latency”. In the public Internet, there’s lots of latency, long and variable, as individual packets of data are delivered via widely varying routes through the network. For a discussion via electronic conferencing, that means all those subtle unwritten rules–about when and how it’s OK, or not OK, to chip in and interrupt–get really disrupted. Users find that very unsettling

When will the public Internet be good enough so that it’s the default for the audio element of conferencing? Our current guess is around 2014 in most rich countries. In other places, probably quite a bit longer than that. Could be 2040 in some places.

However, that only applies to users who depend on the public Internet, with its best efforts/no promises level of Quality of Service (QoS). More and more businesses are buying Wide Area Ethernet services from telephone companies. Those are IP networks, like the public Internet, but unlike the public Internet they use quality-enhancing techniques like MPLS that reduce latency, lost packets and other QoS problems. Users of such services will likely be able to realize the full benefits of routing both the voice and data parts of a conferencing session together over a data network, long before it becomes the usual method for users who rely entirely on the public Internet.

… David Ferris, with thanks to Michael Tyler for his helpful input