Conference PaperPDF Available

Collaboration as a second thought

Authors:
Conference Paper

Collaboration as a second thought

Abstract and Figures

Collaboration is often an afterthought to a project or development. In this paper we describe and analyze our experiences in developing collaborative technologies, most often involving the sharing of visual information. We have often developed these in a context that required us to retrofit existing analysis applications with collaboration capabilities. This approach, though fruitful, is time-consuming, expensive, and often difficult to re-apply elsewhere - it is just hard to change an existing application. One way to make such an effort easier is to package the collaborative components as a kit that can be leveraged on a case-by-case basis. The fixed interface provided by a well-designed toolkit eases the integration process by providing an unchanging and familiar set of components to deploy. Better still, we find, are approaches that require no modification of applications while providing rich and powerful means for sharing information with collaborators. We discuss three separate and illuminating examples that meet this criterion in different ways: (1) building the collaborative potential into the underlying abstractions on which operating systems are built, (2) building tools that live side-by-side with any application in the context provided by the operating system, or by (3) building information tools that use collaborative modalities to better integrate with our workflow. These are probably not the only options, but they all derive from an approach where collaboration is considered early in the design process and therefore manifests itself deep in the computing infrastructure giving it a wider cast.
Content may be subject to copyright.
Collaboration as a Second Thought
Mark Hereld, Michael E. Papka, Thomas D. Uram
Mathematics and Computer Science Division
Argonne National Laboratory
{hereld, papka, turam}@mcs.anl.gov
ABSTRACT
Collaboration is often an afterthought to a project or
development. In this paper we describe and analyze our
experiences in developing collaborative technologies,
most often involving the sharing of visual information.
We have often developed these in a context that required
us to retrofit existing analysis applications with
collaboration capabilities. This approach, though
fruitful, is time-consuming, expensive, and often difficult
to re-apply elsewhere – it is just hard to change an
existing application. One way to make such an effort
easier is to package the collaborative components as a
kit that can be leveraged on a case-by-case basis. The
fixed interface provided by a well-designed toolkit eases
the integration process by providing an unchanging and
familiar set of components to deploy. Better still, we
find, are approaches that require no modification of
applications while providing rich and powerful means
for sharing information with collaborators. We discuss
three separate and illuminating examples that meet this
criterion in different ways: (1) building the collaborative
potential into the underlying abstractions on which
operating systems are built, (2) building tools that live
side-by-side with any application in the context provided
by the operating system, or by (3) building information
tools that use collaborative modalities to better integrate
with our workflow. These are probably not the only
options, but they all derive from an approach where
collaboration is considered early in the design process
and therefore manifests itself deep in the computing
infrastructure giving it a wider cast.
KEYWORDS: Collaboration, Visualization, Best-
Practices, Human-Computer Interaction
1. INTRODUCTION
The 2007 report on Visualization and Knowledge
Discovery [1] concludes that basic research is needed in
several areas to enable continued scientific discovery,
given the current trajectories of data produced by
simulations and collected from sensors. Interaction and
collaboration are at the top of the list:
A new generation of visualization and data
exploration tools are needed to significantly
enhance interaction and collaboration between
these distributed scientists, their data, and their
computational environments.
What is interesting is that this statement clearly
acknowledges three components of a successful
collaboration: scientists, data, and the environment itself.
Building on a long history of creating collaborative
environments to support science, we discuss in this paper
the challenges associated with taking an existing tool and
making it “collaborative.” We present the various
methods we have used, from leveraging tools that make
an application appear to be collaborative, to adding
infrastructure in an ad hoc manner to enable collaborative
use, to building on a toolkit or framework that supports
collaboration. We start with a brief background on the
collaborative spaces we have explored. We then relate our
experiences in retrofitting tools to make them
collaborative. We discuss a few environments and tools
that have collaboration at their very core. From our
experiences we suggest the next steps in supporting
effective collaboration.
2. BACKGROUND
Starting in 1994 we began to experiment with the
construction of collaborative virtual environments. Our
early work progressed from the first CAVE-to-CAVE
application, to the development of the CAVEcomm
library [2] and the advanced ManyWorlds framework [3],
and finally to a desktop system called Metro. At the same
time we were looking at how to leverage efforts in the
community in the construction of middleware to support
distributed workspaces [4].
These experiences made us realize that, while
considerable commercial effort was being devoted to
solving the desktop-to-desktop scenario, certain areas
were still being ignored: (1) group-to-group
collaborations among people at different institutions and
(2) a way to integrate tools people use when meeting face
to face. Video conferencing solutions with single cameras,
click to talk, and a single microphone did not meet such
needs. They were also not very engaging or natural
environments.
To address this situation, we initiated the Access Grid
project [5]. Our objective was to design and build a
deployable tool to enable groups of individuals at remote
locations to collaborate and work as if they were
collocated. The Access Grid is an open-source research
project dedicated to enabling collaboration between
groups of individuals, through the sharing of audio,
video, text, and applications. The Access Grid represents
a scalable solution for collaboration ranging from the
desktop to room-size suites. The development of the
Access Grid toolkit, which is now in version 3.1,
represents significant experience in working with a large
community of diverse users. The Access Grid currently
supports over 2,000 users in over 50 different countries.
Initial prototypes of integration of visualization
infrastructure with the framework have been demonstrated
over the past five years [6,7] and more recently in a
production environment for the volumetric rendering of
medical images [8]. In conjunction with this effort we
have been planning and prototyping community-
accessible visualization solutions to simplify access to
resources [9]. To date, collaboration technology has not
advanced beyond the situation that was the impetus for
the development of the Access Grid.
3. COLLABORATION RETROFITTING
Collaboration has long been an important element of
scientific research; generally this has been a local
collaboration in which collocated colleagues share
scientific and computational resources. As collaborations
become more distributed, it is essential that the capability
continue. In many cases, however, the applications that
scientists rely on are not collaborative and often are not
even network-aware. These applications must be
retrofitted for collaboration so that people in
geographically distributed locations can use them and
have a common sense of the workflow.
3.1. Manual Retrofitting
Multiple approaches to manual retrofitting have been
pursued. Two common approaches are the use of remote
screen buffers and implementation of collaborative
facilities within the application. The most trivial is that
of screen sharing à la virtual network computing (VNC),
in which the content of the display is captured and
transmitted to remote participants, from whom
interactions can be captured and transmitted back to the
source application. The wide applicability of this
solution is appealing, as it is largely application- and
platform-independent. It does not, however, preserve any
shared representation of the application state, essential to
the continuing pursuits of the collaborative endeavor.
VNC also faces well-known security and connectivity
challenges.
A second approach for equipping an application for
collaboration is to implement this functionality directly
in the application code. This approach has its merits,
primary tight integration and full control of the
application. We have used this approach in numerous
instances, for example in writing stream-processing units
for Chromium to stream-rendered frames as video. The
downside of this approach is its labor-intensive nature, in
that it requires not only design of novel collaborative
facilities but also deep knowledge of the target application
code. Having followed this approach in several instances,
we have identified common facilities and patterns,
clarifying the need for a collaboration toolkit.
3.2. Retrofitting with a Collaboration Toolkit
The Access Grid establishes a collaborative environment
with audio, video, text chat, shared data, and shared
applications. Beyond the typical user tools for
collaboration, the Access Grid includes a toolkit for
developing shared applications. The collaborative
components of the toolkit are key to establishing and
maintaining shared distributed state: a centralized store of
shared application state and an event distribution service.
The centralized application state consists of a memory-
resident keyword-value store. The storage interface itself
places minimal constraints on the structure, content, and
format of the data, instead leaving these decisions to the
application.
The event service is similar in its treatment of data: it
includes facilities for addressing and messaging, but
regards the data payload as entirely opaque. Performance
was a key consideration in the design of the event service;
using lightweight data formats, the event service is able
to deliver messages between users distributed around the
planet with minimal overhead. A further benefit of the
event service is that it avoids firewall problems by relying
strictly on outgoing connections from users. These
facilities provide the underpinnings necessary for
applications to be extended for collaboration.
Implementing collaborative facilities in an existing
application is significantly simplified when one need not
be concerned with the format of data or means of
distribution. In the next section, we give examples of
standalone applications that have been extended with
collaborative facilities using the application development
components of the Access Grid.
4. EXPERIENCE
Collaboration can be divided into two functionally
distinct layers: remote presentation and remote
interaction. Remote presentation involves simply sharing
the content of one’s work with remote viewers,
presumably with some out-of-band mechanism for
discussing the content. Thishalf-duplex” interaction is
often an acceptable level of collaboration in situations
involving a single expert or localized data or application
base and a group of remote observers.”
Remote interaction involves bidirectional or “full duplex”
interaction, in which the remote collaborators can fully
control the application being shared. This scenario is an
extension of remote presentation in that when all
collaborating participants have equal expertise with the
application or data being presented, any one of them can
take over and discuss the shared content.
The following sections describe our experience with each
of these mechanisms.
4.1. Streaming ParaView-rendered Data
ParaView is a widely used application for visualization of
scientific datasets (Fig. 1). These datasets are often large
and reside on remote data storage resources, with which
the ParaView client interacts through a built-in client-
server mechanism, such that the server reads the data and
renders it and the client simply displays it.
Scientific visualization is typically undertaken by
visualization experts operating on data that belongs to
scientists. These two groups work together on the
development of the visualization to ensure that it
accurately represents the data, usually over a period of
time proportionate to the size and complexity of the
dataset.
As research groups become increasingly distributed, these
interactions must also become distributed. To model
these interactions in a distributed environment, the
visualization expert and scientist require a mechanism for
reviewing intermediate visualization results and adjusting
the space, time, and color representations from their home
institutions.
The first step in our solution is to allow the visualization
expert to work in the usual fashion, in this case using
ParaView, with an extension that streams the resulting
visualization over the network as video to the remote
scientists. The scientists run an application that can
receive and display the incoming visualization data. The
visualization frames are streamed by using a video codec
that optimizes for bandwidth consumption by limiting
the data it sends according to changes between frames
(e.g., motion coding). This allows the stream to achieve
a high frame rate and low latency, making it a convenient
solution for collaborative visualization (Fig. 2).
Used in conjunction with the Access Grid, which
includes the application that receives and displays the
incoming visualization video, the visualization expert and
the scientist(s) can interact with each other and the
visualization in a fashion similar to being in the same
physical room.
Subsequent development will focus on enabling the
remote scientists to take control over ParaView and adjust
the visualization, as over time the domain scientists will
become more comfortable with the visualization tools and
it will be more natural and faster for them to interact with
the visualization themselves.
4.2. Access Grid Shared Applications
Here we give four examples of shared applications that
have been built with the collaborative components
provided by the Access Grid: Shared Presentation,
Shared Rasmol, Shared Gnuplot, and vl3.
Shared Presentation is a tool for distributed control of
slide presentations. This application uses the data store
for state, such as the location of the set of slides (e.g., a
URL), the current slide number, and the identity of the
presenter (for floor control). The event service is used to
Figure 1. Screen capture of ParaView application
being used to look at data from a simulation of blood
flow within an artery. (Data provided by George
Karniadakis - Brown University)
communicate events such as the loading of a slide set,
advancing to the next slide, and floor control changes.
Rasmol is an application used by biologists for three-
dimensional visualization of molecular structure. We have
built Shared Rasmol to allow biologists to
collaboratively view molecular models from remote
locations. Shared Rasmol opens model files from the
Access Grid Venue datastore and tracks application
interactions for rotating, panning, and zooming the model
and changing the display of the model. These interactions
are communicated to remote users where they are pumped
into the receiving Rasmol instance to effect the same
view. The interactions are communicated over the event
service, which must bear the heavy load of frequent
updates to the model position and transformation
resulting from mouse interactions.
Gnuplot has a long history in graphical viewing of
scientific data by scientists in many domains. Shared
Gnuplot accepts input from any member of a group of
users and displays the resulting plot at all remote sites.
As with other shared applications, this allows one user
with expert knowledge in either the tool (Gnuplot) or the
domain data to share his expertise with the group.
The vl3 tool is a distributed, cluster-based volume
visualization application with a collaborative frontend.
The rendering engine is run on the cluster, and the user
interface communicates changes in the view back to it for
rendering updates. Discovery of the vl3 session occurs
through the Access Grid Venue (as is typical of AG
shared applications). When a user joins the vl3 session,
the client launches against addresses stored in the shared
application state, and interactions with the client
propagate to remote collaborators over the Access Grid
event service.
4.3. Results
The examples above describe two approaches to building
collaborative applications: remote presentation and remote
interaction. A hybrid of these two approaches, best
exemplified in the case of vl3, may be the best approach
going forward. Streaming the relevant application interface
to remote participants as video optimizes for presentation
and bandwidth but precludes interaction. The Access Grid
facilities used by the shared applications described above
arose out of many instances of shared application
development and provide the necessary support for remote
interaction.
5. FUTURE POSSIBILITIES
Interesting environments can arise when one begins by
considering collaboration instead of ending there. In this
section we present three ideas for collaborative
technologies. We think that they create collaborative
experiences that are very different from what we have seen,
even though they rely heavily on ideas that have been
tried in part and in other forms. The three examples
below illustrate powerful collaborative capabilities
available to all applications in conventional personal
computing environments.
5.1. The Vast Pixel Savannah
How can we share what we are seeing on our display
screens in a simple and fluid way? This question lies at
the heart of collaborative visualization.
The need for a new paradigm in handling pixels is being
expressed in several areas of ongoing work. At the more
traditional end of the scale are the solutions that consider
end-point displays as containing graphics horsepower
according to the usual workstation standards. Multi-
display systems using WireGL, Chromium, DMG, and
Figure 2. Photos of users participating in collaborative ParaView session; the user sitting at his desk is sharing
the visualization results with two colleagues at a different location.
Figure 3. The Vast Pixel Savannah, a universally
shared persistent visual space.
similar solutions are of this ilk. A central problem for
architectures based on this paradigm is that operating
system integration with the display adaptor has evolved
to provide powerful graphics pipelines from application to
pixels that are difficult to break into. At the other end of
the scale, solutions that put pixels in the prime position
(as opposed to graphics primitives) are providing perhaps
the most flexible and neutral paradigm, typically at the
expense of performance. SAGE, VNC, and the many
progeny of VNC are among the approaches that fit into
this category. The venerable and ancient X lies between
these extremes.
With the vast pixel savannah (VPS), we consider a model
where pixels are primary—shipped à la carte across the
network—and introduce the notion of a global address
space for pixels that is accessible to all display devices.
Shown in Figure 3 is an example of two conventional
work planes, a laptop and a dual headed workstation, that
are making use of two windows onto the VPS. Each is
windowed onto more or less permanent homestead plots,
which in this case are temporarily connected via a bridge
(labeled wormhole) to facilitate passing windows. Each is
additionally windowed onto a larger shared pixel plot
(labeled playground) that might be the target of a large
visualization that these two collaborators are exploring.
Thrown in for the purpose of illustration are plots in the
VPS corresponding to the display of a cell phone and a
public digital billboard
. Pixel real estate might be apportioned by using DNS-
like services over the Internet. Every display device that
adopts this model would face many of the same
administrative questions that it faces with respect to
network address: where to begin, if and when to change
the position of its view onto the VPS, who to ask if it
doesnt know what to do, and what coordinates are legal.
Taking this view of display space would enable many
useful possibilities. One’s display becomes a window
onto this terrain. One may own a private plot, or pixel
homestead. Bridges between regions create expedient
new relationships by manipulating the topology of the
VPS. An extreme interpretation of the VPS that may
open its use to new and interesting possibilities is as a
kind of collaborative terrain along the lines of Second
Life.
5.2. Pixel Porting
Sometimes collaboration is best served by sharing only a
small piece of a visual representation right now and
without any prior warning. The barrier to such
impromptu visual collaborations is too high for most
people's taste, and so it rarely contributes to the
collaborative workflow.
Standing between collaborators and a fluid flow of shared
visual experience are a host of issues including non-
uniform interfaces to applications, computing platform
differences, the semantics of the desktop metaphor, and
the mechanics of keyboard and mouse. Issues of personal
preference also play a key role in this problem. One can
see instances in existing applications and modes of
human-computer-human interaction that capture useful
styles. Instant messaging has many properties that could
aid us here, the spontaneity afforded by an open channel
and its lightweight interface among them. Keyboard
shortcut tools (Quicksilver comes instantly to mind) can
be extremely powerful; though can present a bewildering
if not daunting array of magical keystrokes that is off-
putting to many. Direct selection of pixels from the user's
field of view is an exceptionally useful tool because it has
the power to represent exactly what interests while
skipping over the encumbrances of the desktop metaphor
focus is on the flat visual imagery.
We believe that a useful tool capable of promoting
frictionless sharing of visual data can be constructed out of
the best aspects of these many components. Here is a
sketch of what we envision, in the form of a use case.
You see something interesting on your screen that you
would like to share with a collaborator. You hold the
Ctrl-Option meta-keys while dragging a selection box
over the area of your screen that contains the important
content. The highlighted region (a pixel portal) can then
be dragged onto your IM client where it resolves to an
existing conversation or creates a new one as
circumstances dictate. In Figure 4, the provider on the
right (you) can resize or move using corner handles. The
widget bar below the portal window contains selectors for
still, new snap, continuous, and remote control. The
widget bar below the destination port (mine), if enabled
by the provider, allows the subscriber to steer the pixel
portal’s mouth end around the window or screen on the
provider’s laptop. Decorating the conversation now is an
active region with the contents of the selected pixel
Figure 4. Sketch of a pixel port connection.
portal; it can be a snapshot of the contents at the moment
of selection or, at your option, a constantly updated
image mirroring your selection. At any time you can
resize or move the pixel portal with instantaneous effect
on what your collaborator sees. You may allow your
collaborator to steer. Pixel portals can be managed by a
mechanism parallel to the ubiquitous clipboard, or they
can be redirected to other or additional collaborators.
With this notion in place, many potentially useful
embellishments are possible.
5.3. IM Everywhere
As our research workflows become increasingly complex
and intertwined, not only with our human collaborators,
but also with the range of computing processes that
facilitate all aspects of our daily chores, there is increasing
need for human-centered approaches in order to relieve the
mental overhead that attends this complexity.
Perhaps it is opportune to start thinking about broadening
the sphere of influence of collaboration research to include
the non-human partners in our work. Human-centered
approaches to computing are increasingly feasible as
technologies in natural languages, agents, security,
ubiquitous computing, and grid-computing (to name a
few) are advanced and integrated into our common
computing fabric.
The idea embodied in IM Everywhere is to deploy our
agents behind the natural interface provided by the IM
client. By packaging our computing aids as agents and
wrapping them in the conversational interface we use for
our human collaborations we reduce the number of
different command and control modes that we are required
to master, reduce the number and depth of context
switches required to manage our multi-tasking, and begin
an evolutionary process that will ultimately enable us to
focus more exclusively on our research problems and less
on the problems created by the technologies we hurl at
our research problems.
6. CONCLUSIONS
Current collaborative infrastructure available today is not
enough; we need collaborative services that integrate with
scientific workflow and data management systems.
Systems must support and promote remote and
collaborative visualization and have algorithm and
infrastructure optimizations to make them usable and
robust. Collaboration technologies available today largely
target videoconferencing, webcasting, or shared
whiteboard technologies and are expensive, closed
systems often using proprietary protocols.
We advocate an approach that is based on open source
infrastructure designed for easy integration into a wide
variety of applications and environments. With this
approach, even simple ideas could provide powerful new
modes of collaboration that fit invisibly into the working
environment and style of the research scientist. We have
described three separate ideas that fit this model: (1) the
Vast Pixel Savannah the collaborative potential is built
into the underlying abstractions applying to our display
surfaces, (2) Pixel Porting relies on a ubiquitous tool
that lives side-by-side with any application, and (3) IM
Everywhere conveys collaborator status to even our
most mundane of information sources by wrapping them
in agent technology and opening an instant messaging
channel to them. These are probably not the only
options, but they are all designed into the working fabric
of our computing environments well below (or outside of)
the imaginary box containing an application. This
property endows them with the power of independence
making them always available.
ACKNOWLEDGEMENT
This work was supported by the Mathematical,
Information, and Computational Sciences Division
subprogram of the Office of Advanced Scientific
Computing Research, Office of Science, U.S. Dept. of
Energy, under Contract DE-AC02-06CH11357.
REFERENCES
[1] Johnson, C., R. Ross, S. Ahern, J. Ahrens, W. Bethel, K.-L.
Ma, M. Papka, J. v. Rosendale, H.-W. Shen, and J. Thomas,
"Visualization and Knowledge Discovery: Report from the
DOE/ASCR Workshop on Visual Analysis and Data
Exploration at Extreme Scale," Salt Lake City October
2007.
[2] Disz, T. L., M. E. Papka, M. Pellegrino, and M. Szymanski,
"CAVEcomm Users Manual," Argonne National
Laboratory, Argonne, Technical Memoramum ANL/MCS-
TM-218, September 1996.
[3] Disz, T. L., R. Olson, M. E. Papka, R. Stevens, M.
Szymanski, and R. J. Firby, "Two Implementations of
Shared Virtual Space Environments," Argonne National
Laboratory, Argonne, Preprint ANL/MCS-P652-0297,
March 1997.
[4] Foster, I., M. E. Papka, and R. Stevens, "Tools for
Distributed Collaborative Environments: A Research
Agenda," in High Performance Distributed Computing
Focus Workshop on Multimedia and Collaborative
Environments,, Portland, Oregon, 1996.
[5] Stevens, R., "Group-Oriented Collaboration: The Access
Grid Collaboration System," in The Grid 2: Blueprint for a
New Computing Infrastructure, 2nd Edition ed, I. Foster
and C. Kesselman, Eds.: Morgan Kaufmann, 2004, pp.
191-200.
[6] Olson, R., and M. E. Papka, "Remote Visualization with
VIC/VTK," in Visualization 2000 Hot Topics Salt Lake
City, UT: IEEE Computer Society, 2000.
[7] Karonis N., M. E. Papka, J. Binns, J. Bresnahan, J. Insley,
D. Jones, and J. Link, "High-Resolution Remote Rendering
of Large Datasets in a Collaborative Environment," Future
Generation of Computer Systems, 2003, pp. 909-917.
[8] Binns, J., F. Dech, M. E. Papka, J. C. Silverstein, and R.
Stevens, "Developing a Distributed Collaborative
Radiological Visualization Application," in From Grid to
HealthGrid, 2005, pp. 70-79.
[9] Binns, J., J. DiCarlo, J. A. Insley, T. Leggett, C.
Lueninghoener, J.-P. Navarro, and M. E. Papka, "Enabling
Community Access to TeraGrid Visualization Resources,"
Concurrency and Computation: Practice and Experience,
vol. 19, 2006, pp. 783 - 794.
ResearchGate has not been able to resolve any citations for this publication.
Article
Full-text available
The CAVEcomm library is a set of routines designed to generalize the communications between virtual environments and supercomputers.
Article
Full-text available
While many issues in the area of virtual reality (VR) researchhave been addressed in recentyears, the constant leaps forward in technology continue to push the field forward. VR research no longer is focused only on computer graphics, but instead has become even more interdisciplinary,combining the fields of networking, distributed computing, and even artificial intelligence. In this article we discuss some of the issues associated with distributed, collaborative virtual reality,aswell as lessons learned during the developmentoftwo distributed virtual reality applications. 1 Introduction The Futures Laboratory at Argonne National Laboratory has been exploring what is needed to support large-scale shared space virtual environments (VE) for wide-area collaborations. Our research has focused on the system architecture, software design, and features needed to implement suchenvironments. In this article we discuss twoprototype systems under development at Argonne. Shared virtual spaces a...
Chapter
This chapter introduces an important class of Grid applications—the Access Grid system— aimed at supporting wide-area, real- time, computer-mediated communications, The Access Grid project began in 1994 to develop collaboration environments based on two assumptions: (1) over time, the amount of bandwidth and computing available to research groups would become virtually limitless; and (2) most scientific collaborations involve groups of people at each collaborating institution, not just single individuals. In designing the first Access Grid (AG), the focus was on basic enabling infrastructure, using open source software wherever possible, avoiding commitment to a particular vendor, and encouraging participants to contribute to the research and development of AG technology. The vision of the Access Grid reflects the belief that, within the near future, bandwidth, computing, and imaging power may become effectively free and that high-quality audio and video capture may be increasingly inexpensive. The challenge may associated with the ways to organize these capabilities to support high-end scientific work—to create environments that people want to spend time in and that encourage experimentation and growth.
Article
In a time when computational and data resources are distributed around the globe, users need to interact with these resources and each other easily and efficient. The Grid, by definition, represents a connection of distributed resources that can be used regardless of the user’s location. We have built a prototype visualization system using the Globus Toolkit, MPICH-G2, and the Access Grid in order to explore how future scientific collaborations may occur over the Grid. We describe our experience in demonstrating our system at iGrid2002, where the United States and the Netherlands were connected via a high-latency, high-bandwidth network. In particular, we focus on issues related to a Grid-based application that couples a collaboration component (including a user interface to the Access Grid) with a high-resolution remote rendering component.
Conference Paper
Argues that future computing environments will be collaboration-oriented, globally distributed and computation/information-rich. These environments will be accessed via multiple interface devices. As we move from a desktop-centric computing model to a network-centric model, new approaches in the way software and data are handled will need to be developed. In this article, we outline the requirements for enabling the technological infrastructure, describe some first steps that we have taken toward building this infrastructure, and sketch directions for future development.
Article
Visualization is an important part of the data analysis process. Many researchers, however, do not have access to the resources required to do visualization effectively for large datasets. This problem is illustrated through several user scenarios. To remedy this problem, we propose a Visualization Gateway that provides simplified access to such resources to a broad population of users. The current implementation of this gateway is described, including technology used and services made available. In particular, a detailed description of a ParaView portlet is included. A proposed design for enabling access to community users is discussed. Technology as well as policy issues that were raised, including security and data management, are covered. Methods for providing additional services, scaling to include additional resources, and other areas of future development are conveyed. The paper concludes with a summary of the topics covered.
Leveraging the advances of today's commodity graphics hardware, adoption of community proven collaboration technology, and the use of standard Web and Grid technologies a flexible system is designed to enable the construction of a distributed collaborative radiological visualization application. The system builds from a prototype application as well as requirements gathered from users. Finally constraints on the system are evaluated to complete the design process.
Visualization and Knowledge Discovery: Report from the DOE/ASCR Workshop on Visual Analysis and Data Exploration at Extreme Scale
  • M Ma
  • J V Papka
  • H.-W Rosendale
  • J Shen
  • Thomas
Ma, M. Papka, J. v. Rosendale, H.-W. Shen, and J. Thomas, "Visualization and Knowledge Discovery: Report from the DOE/ASCR Workshop on Visual Analysis and Data Exploration at Extreme Scale," Salt Lake City October 2007.
Remote Visualization with VIC/VTK
  • R Olson
  • M E Papka
Olson, R., and M. E. Papka, "Remote Visualization with VIC/VTK," in Visualization 2000 Hot Topics Salt Lake City, UT: IEEE Computer Society, 2000.