This document classifies and discusses different application sharing tools that could be used in a virtual classroom setting. It examines tools based on their architecture (centralized vs replicated), how application sharing is implemented in conferencing systems, thin-client systems, and peer-to-peer networks. It also provides a table comparing several popular application sharing tools based on factors like the protocol used, ability to change presenters, support for multiple users, and whether it allows remote control. The document concludes by noting that centralized application sharing requires high bandwidth and low latency to perform effectively.

1. A Classification of Application Sharing Tools for Use
in a Virtual Classroom
Gerhold B. Kooper1
, Prof. Alfredo Terzoli2
Computer Science Department, Rhodes University, South Africa
Tel: +27 46 6038247, Fax: +27 46 6361915
email: gbkooper@gmail.com1;
A.Terzoli@mail.ru.ac.za2
Abstract – In this paper we classify different
application sharing tools according to factors that
would affect their use in a virtual classroom. We
exemplify this classification with a subset of
different types of application sharing tools, such
as conferencing systems, thin-client systems and
peer-to-peer systems. We also discuss
performance issues relating to application sharing
tools.
Index Terms - Application Sharing, Desktop
Sharing, Virtual Classroom, Collaboration
I. INTRODUCTION
The Computer Science department at Rhodes
University provides distance courses to the Computer
Science department of the University of Namibia, to
help with staff and research capacity building there.
The teaching takes place over the Internet in a virtual
classroom.
A. The Virtual Classroom
As stated in [10], “A virtual classroom should not
be much different from a real classroom”. In a virtual
classroom one would like to have the teaching and
learning experience as close as it would be in the real
classroom as possible. The difference from a real
classroom should be only that a “virtual place” is
created, which allows teachers and students to
connect over any distance and from any geographical
location.
Such a virtual classroom would be a very efficient
and practical way of utilizing human resources in a
university environment. Lecturers would then not
need to prepare and present different lectures for the
physical and the virtual classroom.
Within a virtual classroom there is a need for
computer application sharing, for example to allow
lecturers to run experiments with software.
In an African context, the choice of application
sharing technique must take into account low
bandwidth and unstable network connections.
B. Application Sharing
From the point of view of a virtual classroom,
application sharing can be described as the sharing of
the graphical user interface of an application amongst
multiple users simultaneously in real-time. All
participants have the same view of the application.
The whole desktop might be shared rather than a
single application, in which case one speaks of
desktop or screen sharing. Application sharing and
desktop sharing will be considered equivalent in this
paper except where the difference is explicitly made.
In this paper “application sharing software” is the
software that handles and connects the different
participating computers to enable the sharing.
II. DESIGN ARCHITECTURE
An application sharing system can be categorized
in terms of how it handles the shared application,
namely centralized or replicated [6].
In the centralized architecture there is only one
instance of the shared application, whereas the
replicated architecture requires each participant to run
locally his/her own copy of the shared application.
The replicated architecture, by its design, may be
very bandwidth efficient and faster than the
centralized architecture as only input streams are sent
among participating computers. It was even stated by
C. Ming and his team at the HP Software Technology
Laboratories that they could share a 3D application
over just 56kbps network connection [2].

2. The replicated architecture has however not
survived in application sharing software over the
years and we could not find usable software solutions
using this architecture. It suffered greatly from the
problem of keeping all shared application instances
synchronized [5].
Replication architectures seem to be successfully
applied in networked gaming. A gaming session may
sometimes have more than tens of thousands of
concurrent players. In these games, only the actions
of the players would be distributed amongst players
rather than the graphics updates [11].
III. TYPE OF APPLICATION SHARING SYSTEMS
We can further classify application sharing
software in respect to the systems in which they are
used. In the subset of application sharing tools
presented in Table 1, all the tools use the centralized
architecture, but they are different along other
dimensions.
C. Application Sharing within Conferencing
Systems
In many instances, application sharing would be
part of a video or voice conferencing system. The
sharing of applications in such systems would be
referred to as data conferencing or desktop
conferencing. Examples in Table 1 are Citrix
GoToTraining, Cisco WebeX, Microsoft Live
Meeting and Adobe Acrobat CONNETNOW.
These systems generally have a good mechanism
of floor control to handle changes to the application
being shared [3].
D. Application Sharing within Thin-Client
Computing Systems
A thin-client computing system consists of a server
and a client that communicate over a network using a
remote display protocol. Thin client systems are
increasing becoming popular over the internet with
the increasing presence of technologies such as cloud
computing.
Application sharing can be done with thin-client
systems although these systems are not primarily
designed for desktop conferencing. With appropriate
configuration, however, they can form a desktop
conference session. Examples from Table 1 are
FreeNX and TightVNC.
A desktop sharing conference was done with the
FreeNX server on Ubuntu Linux by allowing
participants to attach to an existing desktop display.
Three participants shared one screen in this instance.
Sharing of the same display when using TightVNC
can be achieved, for example, with the use of
software called VNC Reflector that allows
broadcasting of a display.
The limitation of these systems would typically be
the lack of a built in mechanism for floor control.
Floor control would have to be done through voice
communication or another communication channel
active amongst participants.
E. Application Sharing within Peer-to-Peer
Systems
In its pure sense, peer-to-peer (P2P) networks or
computing is a distributed system without any
centralized control. Nodes act as both servers and
clients. Hybrid P2P systems on the other hand have a
central server that maintains; for example,
information about registered users on the network, in
the form of meta-data [1]. Many of the modern P2P
systems are designed as hybrid P2P systems.
In the sample software in Table 1 we can regard
Skype and MSN Messenger as Hybrid P2P software
and they allow application or desktop sharing.
IV. SOFTWARE SOLUTIONS
A list of application sharing solutions was
compiled by gathering as much information about
these software solutions through the web and specific
trials.
In table 1 we present a subset of 10 application
sharing solutions from more than 50 solutions that
were investigated. We have selected this subset
because of the popularity and availability of these
solutions, together with their diversity.

3. Notes:
* These solutions would attempt to use a direct connection. If it fails due to network restrictions, data would be relayed through
an intermediate entity.
~ It could not be determined.
The columns in this table are described below.
A. Protocol
The “protocol” column indicates application level
protocol used by the solution.
The ITU-T T.120 family of protocols is the only
standard for application sharing protocols. Some of
the other applications sharing protocols, even the
proprietary ones, were developed by adding their
own adjustments to the T.120family.
It is popular for newer solutions to use the HTTP
protocol to tunnel their proprietary protocols, as is
shown in the subset of solutions presented in Table 1.
This allows their solutions to work on different
platforms and networks. Most universities and other
corporations have network restrictions to protect their
networks and may only allow HTTP internet traffic
through their networks.
B. Relay
The “relay” column indicates whether the solution
uses a central server to relay its screen update data.
C. Remote Control
The “remote control” column indicates whether the
solution would allow for remote participants to
control the hosted application.
D. Share Specific Application
This column indicates if the solution allows only
selected applications to be visible to others in the
application sharing session.
E. Change Presenter
The “change presenter” column indicates whether
the solution allows the “presenter” role to be handed
over to another participant. (A “presenter” in this
context is the person in control of the application.)
F. Live Cursor
The “live cursor” column indicates whether every
movement of the cursors would be visible on all
participating computers in real-time.
G. Multiple Users
The “multiple users” column indicates whether the
solution allows more than two participants to access
the same shared application at the same time.
H. TYPE
The “type” column indicates in which type of
system the application sharing is used as discussed
earlier.
Application Protocol Relay Remote
Control
Share
Specific
Application
Change
Presenter
Live
Cursor
Multiple
Users
Type
Citrix GoToTraining Proprietary over HTTP Yes Yes Yes Yes Yes Yes Conferencing
Cisco WebeX Proprietary over HTTP Yes Yes No No No Yes Conferencing
Microsoft Live Meeting Proprietary over HTTP Yes Yes ~ ~ Yes Yes Conferencing
Adobe Acrobat
CONNETNOW
Proprietary over HTTP Yes Yes No No Yes Yes Conferencing
TeamViewer Proprietary over HTTP No* Yes Yes No Yes ~ Conferencing
Mikogo Proprietary over HTTP Yes Yes No Yes Yes Yes Conferencing
Skype Proprietary over HTTP No* Yes Yes No Yes ~ P2P - Hybrid
MSN Messenger MSN Messenger Protocol Yes Yes No No Yes ~ P2P - Hybrid
FreeNX NX No Yes No ~ Yes Yes Thin-Client
TightVNC VNC/RFB No Yes No ~ Yes Yes Thin-Client
TABLE 1: APPLICATION SHARING SOLUTIONS

4. V. PERFORMANCE ISSUES
Centralized application sharing needs both high
bandwidth and low network latency to perform
effectively. The display area on the screen needs to
be typically bigger than one would see for example in
a Skype video call in order to read the text on the
screen. This would generate large amounts of data to
be transmitted over the network.
In terms of the end points, there are settings and
configurations of the application sharing software
that directly affect the performance. Important ones
are the screen resolution, the colour depth, the
display encoding, the compression algorithms used,
and the cache settings.
The following factors measure the performance of
the application sharing software in terms of the
experience by the participants when using application
sharing software.
A. Latency
Latency is the delay between the moment in which
the presenter initiates an activity, such as mouse
clicks or cursors move on the screen, and the moment
in which the other participants observe the same
action on their computers [8].
B. Smoothness
A smooth experience is one that flows easily with
a minimum of “choppiness” and fuzziness of images
[8]. Smoothness may improve focus of participants
and reduce distraction.
C. Fidelity
Fidelity denotes how accurate a copy is to its source.
When doing application sharing, presentations with
perfect fidelity looks identical to the presenter and
the other participants. With a good fidelity, the
classroom session experience is improved, reducing
the risk of participants to misread and misinterpret.
VI. DISCUSSION
In our virtual classroom it would be preferable to
use a solution with the following characteristics:
 Support of multiple users as the
classroom might have students in
different locations.
 It would not be good for the solution to
use relays. The servers used for relay by
most tools are very often on other
continents such as Europe or North
America and they introduce high network
latency.
 A solution within a conferencing system
would be advantageous as these systems
have built in floor control.
In terms of the performance of the application
sharing solution chosen, the latency seems most
important. Fidelity and smoothness may not affect
the activity in the virtual classroom significantly, at
least until the transmitted session degrades
substantially.
VII. CONCLUSION
In this paper we have presented an initial
application sharing classification, with a view of
providing the right tool for use in virtual classroom
activity. The classification was along dimensions
such as the type of systems, the use of relay servers,
the type of protocol and others.
We have also classified a subset of application
sharing solutions currently in use according to the
criteria listed.
Finally we have also discussed the application
sharing solutions in terms of usability and experience
by users.
VIII. REFERENCES
[1] Pourebrahimi B., Bertels K, Vassiliadis S.
2005. A Survey of Peer to Peer Networks,
Proceedings of the 16th
Annual Workshop on
Circuits. Systems and Signal Processing, Veldhoven.
Netherlands.

5. [2] Hao M. C., Lee D., and Sventek J. S. 1996. A
light-weight application sharing infrastructure for
graphics intensive applications. In Proceedings of the
5th IEEE International Symposium on High
Performance Distributed Computing (HPDC '96).
IEEE Computer Society, Washington, DC, USA
[3] Ahuja S. R., Ensor J. R., Lucco S. E. 1990. A
comparison of application sharing mechanisms in
real-time desktop conferencing systems. SIGOIS
Bull. 11, 2-3 (March 1990), 238-248
[4] Rohall, S. L., Patterson, J. 2004. Another Look
at Replicated-Application Sharing. Position paper for
CSCW 04 workshop on Making Application Sharing
Easy: Architectural Issues for Collaboration
Transparency (Chicago, USA, Nov. 6-10). ACM,
New York, NY.
[5] Rohall Steven L., Patterson John F. 2005. The
Zipper System for Flexible Replicated Application
Sharing, UIST ’05, October 23 – 27, Seattle,
Washington. USA.
[6] Crowley Terrence, Milazzo Paul, Baker Ellie,
Forsdick Harry, Tomlinson Raymond. 1990.
MMConf: an infrastructure for building shared
multimedia applications. In Proceedings of the 1990
ACM conference on Computer-supported
cooperative work (CSCW '90). ACM, New York,
NY, USA, 329-342
[7] Nieh, J., Yang, S. J., and Novik, N. 2000. A
comparison of thin-client computing architectures.
Tech. Rep. CUCS-022-00, Department of Computer
Science, Columbia University, November
[8] Petersen Heather. 2007. Screen Sharing
Performance of Web Conferencing Services:
Competitive Analysis Prepared for Adobe Systems.
White Paper. Handle Consulting. Available at:
http://www.adobe.com/products/acrobatconnectpro/p
roductinfo/features/turbo_screensharing/turbo_screen
sharing_wp.pdf. [Accessed: May 2011]
[9] Lennox J., Schulzrinne H., Nieh J., Barrato R.,
2004. Protocols for Application and Desktop
Sharing, Columbia University. Technical Paper.
Available at: http://
www.cs.columbia.edu/~lennox/draft-lennox-avt-app-
sharing-00.html. [Accessed: April 2011]
[10] Petterson R., Van Limbeek C. A. 2002.
Design of a Virtual Classroom. The 34th Annual
Conference of the International Visual Literacy
Association. October 1 – 5. Breckenridge. Colorado,
USA.
[11] Rhalibi Abdennour El, Merabti Madjid. 2005.
Agents-based modeling for a peer-to-peer MMOG
architecture. Computers in Entertainment (CIE).
Volume 3, Issue 2 (April 2005). New York. USA.
[12] A Brief Look At NoMachine NX [online].
2004. Available at: http://
people.gnome.org/~markmc/a-look-at-nomachine-
nx.html. [Accessed: February 2011]
[13] 30 Amazing Remote Desktop Sharing and
Screen sharing applications [online]. 2010. Available
at: http:// www.developersnippets.com/.../30-
amazing-remote-desktop-sharing-and- screen-
sharing-applications/ [Accessed: April 2011]
[14] Screen Sharing Tools and Technology: A
Mini-Guide [online]. 2006. Available at:
http://www.kolabora.com/news/2006/10/05/screen_s
haring_tools_and_technology.htm. [Accessed May
2011]
[15] Online Training Software & Virtual
Classroom [online]. 2011. Available at:
http://www.gotomeeting.com/fec/training/online_trai
ning. [Accessed: May 2011]
[16] VNC Reflector [online]. 2011. Available at:
http://sourceforge.net/projects/vnc-reflector/.
[Accessed: May 2011]
[17] A SANTED Funded University Collaboration
[online]. 2007. Available at:
http://www.coe.ru.ac.za/VirtualClassroom/Technolog
y/Technology.htm. [Accessed: May 2011]
[18] MSN Messenger Service 1.0 Protocol
[online], 1999,
http://www.hypothetic.org/docs/msn/ietf_draft.txt,
[Accessed: May 2011]
[19] Pallof R. M., Pratt K. 2007. Building online
learning communities: Effective strategies for the
virtual. San-Francisco: Jossey-Bass. John Wiley and
Sons.

6. Gerhold B. Kooper is currently studying towards his
Master of Science Degree in Computer Science at
Rhodes University, Grahamstown. His main area of
interest include application sharing in the virtual
classroom.
Alfredo Terzoli is Professor of Computer Science at
Rhodes University, where he heads the Telkom
Centre of Excellence in Distributed Multimedia. He
is also Research Director of the Telkom Centre of
Excellence in ICT for Development at the University
of Fort Hare. His main areas of academic interest are
converged telecommunication networks and ICT for
development