SEMINAR REPORT

Topic: GOOGLE WAVE

SUBMITTED TO: SUBMITTED BY:

SUBMITTED ON:
 INTRODUCTION
A recent trend in developing innovative eLearning environments is to
combine or create mashups of existing software components,
applications and services. Such an approach increases possibilities for
widening the set of learning scenarios and allows high degree of
personalization. Also, more attention is concentrated on architectures
for supporting personalized learning: through empowering the
functionality of given eLearning environment (integrated approach –
e.g. Elgg) or using individual applications (distributed approach – e.g.
start pages, feed readers, wikis, blogs, etc.) [1], [2], [3], [4]. From
pedagogical point of view eLearning environments are collaborative
and non-collaborative depending upon the way knowledge is created
either through collaborative learning activities or through performing
stand-alone tasks. These environments are controlled, when educator
or institution controls the resources and learning paths and non-
controlled, when learner self-arranges his/her learning environment. In
the blogosphere is started discussion about the difference between
personalized learning and personal learning. Personalized learning is
recognized as controlled and tailored approach where the educator
keeps control over subject area and directs students do the right
learning content, while personal learning is related to self-organized
and life-long learners, e.g. non-controlled approach [5].

When the eLearning environment is well organized and flexible it can

support a range of different interactive learning approaches. It helps in
enhancing the personalized learning. However, it creates different, and
more diverse, demands on the design of eLearning space to support a
range of learning activities, to facilitate group learning approach, to
maximize the use of shared spaces and knowledge. The Google Wave
platform emerged one year ago and has attracted the attention of
educators and researchers with its capability to support users to
communicate and collaborate more effectively on the web in form of
waves [6], [7]. It also supports wiki like collaborative media document
creation. It is based upon a rich set of open standard APIs that allow
integration of extensions and robots within waves that enhance the
base functionality of the Waves and allow embedding of waves in other
web services. The paper explores the characteristics of Google Wave
for personalized learning and they are examined with aim to facilitate
and enhance the knowledge receiving during several engineering
courses and students’ projects. The findings are generalized in a model
that is applied in practice. Students’ opinions are gathered and
summarized.
 CONTENTS

Google Wave is a web-based application that represents a rethinking of

electronic communication. E-mail is 40 years old, predating most of the
technology that people today take for granted, and the basic model of
e-mail remains unchanged. Other forms of electronic communication
have emerged, such as instant messaging, chats, blogs, and texting, and
many communication tools have also migrated to the cloud rather than
running on local campus servers. With these trends in mind, Google is
developing an application that has elements of existing communication
tools but is built around a different model of how communication—and
collaboration—take place. With Wave, users create online spaces called
“waves,” which may include multiple discrete messages and
components—“blips”—that constitute a running, conversational
document. Users access waves through the web, resulting in a model of
communication in which separate copies of multiple messages are not
sent to different people; instead, the content resides in a single space.
People go to a wave to access the con- tent, respond to it, change it,
replay it, send it to a blog, or add new material or attachments.

Google Wave Personalized Learning Settings

Google Wave is designed as a real-time communication and
collaboration platform combining features of email, instant messaging,
wikis, web chat, social networking, and project management. Access
from web browsers and mobile platforms (iPhone, Android) is available
This allows personal/personalized access to the Google Wave from
everywhere and on-demand. The Google Wave platform is extensible
through public APIs, gadgets and robots that allows flexible
personal/personalized content organization. The proposed data model
provides learning in waves with unique IDs. One wave includes
wavelets with unique IDs, each one with a participant list and with a set
of documents. A participant can be a user, a group or even a robot.
Such data model is suitable for knowledge organization in different
waves, performing selective filtering the waves, and creation of special
waves according to student’s needs. Also, personalized learning
happens in collaboration with other learners, working together in a
group learning environment or on a project. The real-time learning
occurs through Wave’s technology known as “operational
transformations”. It allows immediately display of what student types
within a wave when he/she edits a collaborative document used
simultaneously by several other students. The editing operation is also
send to the server to be ratified hoping that it will be accepted by the
server. The server stores all the course documents and all the changes
within the documents. It also displays the latest version of these
documents. In the end, each client is updated with the final version
received from the server that is the result of possibly many operational
transformations. For providing inter/intra-institutional or inter/intra-
course personalized learning, the Google Wave federation protocol can
be used. It allows multiple wave providers to share waves with each
other through open source XMPP protocol. This is a good opportunity
to promote social interactions among students and educators from
different universities and training organizations as well as personal
networking in the context of teaching and learning. A literature
exploration about what is important for personalized learning, what
makes learning personalized is performed below and it is connected to
Google Wave characteristics to support the development of a model for
learning personalization in Google Wave. Research shows that ensuring
of flexible delivery of personalized learning is a main factor guarantees
students’ progression and widening participation according to Different
solutions for personalized way of content delivery are presented in
where a system proposes personalized delivery of news, TV -on-
demand, mobile multimedia applications explores the evolution of the
delivery of content to distributed users in the point of view of a new
level of interaction and personalization providing of today’s Web sites;
in a model of dynamic content delivery supporting a level of
customization and content personalization is presented. Other
important factor recognized by managers, teachers and learners in
supporting the personalization agenda is content and knowledge
presentation in multiple media forms utilizing a wide range of
technologies Technologies allow mashup media content from different
sources and easy embedding of components/widgets for receiving
students’ opinion, practicing basic-skills and assessment, practicing
independent work skills, creating to-do list, pretesting students'
knowledge before each unit, sharing the work of created lessons A wide
variety of media content presentation and knowledge presentation in
different forms attract the attention to learning not only of the
excellent students, but also of excluded/disengaged giving them
alternative routes to access information and skills relevant to their
needs and interests The social nature of web and web applications is a
fact that cannot be skipped. This phenomenon offers opportunities for
personalized learning receipting knowledge by students in formal and
informal situations and scenarios. Working on projects, participating in
forums and group activities are among more formal forms for learning.
Active personalized learning experience can be reached also through
informal conversation, reflexive dialogue and collaborative content
generation, enabling access to a wide raft of ideas and representations
Community of practices is other way for improvement of the
knowledge of each participant through communication and the
possibility for learning through shared experiences, problems and
solutions, tools, methodologies Ubiquities access of web sites and
systems is also influencing factor at receiving a personalized learning
on-demand. There are many examples at using of ubiquities and mobile
environments in the context of personalized learning, as well as
solutions of big companies like Motorola and HP. In the Motorola's
report a model of Seamless mobility ensuring personalization of user
interaction, applications, services and content is presented. It puts the
user at the center of unique experiences remain consistent and
coherent across activities, devices, services, locations and networks The
HP service delivery platform has been designed to delivery of
personalized, content-rich services satisfying specific customer
requirements As it is seen these explored four factors play influential
role for personalized learning organization. They are in scope when the
characteristics for personalized

agenda in Google Wave are examined and when a model for

personalized learning is developed (Figure 1). Flexible delivery is
ensuring by given delivery methods of learning content and knowledge,
technology is represented in support of the methods for an array of
learning resources presentation, learning scenarios are related to the
methods for knowledge reception and mobile access facilitates the
methods for receiving ubiquities/pervasive on-demand learning.
• Delivery methods. Learning instructions, content and knowledge can
reach students through the medium of integrated Google Wave’s rich
text editor, extensions, gadgets and robots. A wave can adopt any
widgets specially created for this purpose, but Wave gadgets are
preferred, because the students can take the advantages of the Wave’s
live, multi-user environment. The Wave robots that are designed as
automated participants within a wave can talk with participants, can
provide information from outside sources, can monitor content within
a wave, can share social bookmarks, etc.

• Content/knowledge presentation. Every student has his/her favorite

way/tools for learning – learning through following links, through well-
formatted text, through multimedia/audio/video/images or embedded
HTML code, some prefer to learn by example, others by finding answers
to questions, and others by solving problems on their own. In a wave all
of these variants are available and students can personalize their
learning using preferably media content. For this purpose they have to
add appropriate gadgets, robots or HTML code.

• Knowledge reception. Google Wave proposes several methods for

knowledge reception – by promoting collaborative work on projects,
small groups’ problem solving, allowing forming virtual learning
network, facilitating learning by others using Google Wave’s playback
function, sharing, searching, real time video conversations, mind
mapping of ideas, discussion through active commenting, and use of
polling gadgets for voting to assess consensus development.

• Methods for ubiquities access. The platform of Google Wave is

reachable any time, from any location, serving the personalized
learning on-demand, including through mobile devices.

Terminology

- Wave: A wave, specifically, refers to a specific threaded conversation.

It can include just one person, or it can include a group of users or even
robots (explained below). The best comparison I can make is that it’s
like your entire instant messaging (IM) history with someone. Anything
you’ve ever discussed in a single chat or conversation is a wave.

- Wavelet: A wavelet is also a threaded conversation, but only a subset

of a larger conversation (or a wave). It’s like a single IM conversation –
a small part of a larger conversation and a larger history. Wavelets,
though, can be created and managed separately from a wave.

- Blip: Even smaller than a Wavelet, a Blip is a single, individual

message. It’s like a single line of an IM conversation. Blips can have
other blips attached to them, called children. In addition, blips can
either be published or unpublished (once again, it’s sort of like typing
out an IM message but not yet sending it).

- Document: A document actually refers to the content within a blip.

This seems to refer to the actual characters, words, and files associated
with a blip.

- Extension: An extension is a mini-application that works within a

wave. So these are the apps you can play with while using Wave. There
are two main types of extenisons: Gadgets and Robots

- Gadgets: A gadget is an application users can participate with, many

of which are built on Google’s OpenSocial platform. A good comparison
would be iGoogle gadgets or Facebook applications.

- Robots: Robots are an automated participant within a wave. They can

talk with users and interact with waves. They can provide information
from outside sources (i.e. Twitter) or they can check content within a
wave and perform actions based on them (i.e. provide you a stock
quote if a stock name is mentioned).
- Embeded Wave: An embeded wave is a way to take a Google Wave
and the conversation within it and place it on your website. Users could
use this as a chatroom, as a way to contact you, or for something more

Wave Gadgets

A Wave Gadget is one of two types of Google Wave extensions.

Gadgets are fully-functional applications. According to Google, gadgets
are primarily for changing the look and feel of waves, although this
seems to only scratch the surface of the potential of a wave gadget.

First: almost any iGoogle or OpenSocial gadget can run within Google
Wave. That means thousands of applications that have been already
created will work in Google Wave. Second: a gadget built within Google
Wave can take advantage of live interaction with multiple users. This
means something like a live online game with active participation from
all users. In that way, it has similarities to Facebook or MySpace
applications, which take advantage of your friend network to make
games, quizzes, and applications more meaningufl and useful.

Gadgets are specific to individual waves, rather than to specific users.

Thus, it’s not like having a Facebook app on your profile – the gadget
belongs to everyone within the wave. They also do not have titles, to
better integrate with the actual conversation. Some of the gadgets
already built include a Sudoku gadget, Bidder (which turns your wave
into an auction), and Maps (which allows for collaboration on a Google
Map).

Wave Robots

Robots are the other type of Google Wave extension. Robots are like
having another person within a Google Wave conversation, except that
they’re automated. They’re a lot like the old IM bots of the past,
although far more robust. Robots can modify information in waves,
interact with users, communicate with others waves, and pull
information from outside sources.

Because it acts like a user, you can define its behavior based on what
happens in the chat. You could build one as simple as “change the word
dog to the word cat” or one as complex as a fully-functional debugger.
We’ll probably start seeming some very advanced robots in the near
future.

Some of the robots already in service include Debuggy (an in-wave

debugger), Stocky (which pulls stock prices based on stock quote
mentions), and Tweety (the Twave robot, which displays tweets inside
of a wave).

Wave Embeds
Wave embeds are a little more complex than embedding a YouTube
video onto your blog, yet in the end, that’s really what Google Wave
Embeds are: a way to take Google Waves onto a third party website.
Embedded Waves support many of the functions of the actual Google
Wave client, including dragging-and-dropping files.

While the Wave Embeds is still very early stage, Google has already
built two: YouTube Playlist Discuss and Multiple Extensions Embed. The
former allows you to discuss a YouTube video via a wave and the latter
allows for interaction with multiple waves on the same page.

One possibility: Google Wave Embeds may be a real-time replacement

to static comments. If Google perfects wave embeds, you could even
see YouTube.com comments replaced with waves, although it is way
too early to make any calls on the potential of this

Furthering your Google Wave education

The Google Wave Logo Still can’t get enough of Google Wave? While
information is sparse, hopefullythis collection of links will help you
understand this new product even more.
 - Mashable’s Google Wave Coverage: We highly suggest bookmarking
our Google Wave coverage and checking Mashable consistently for the
latest information on Google Wave.

- Google Wave Federation Protocol: Google has provided some

community principles, architecture information, and more detailed
definitions on their Wave protocol webpage.

- Google Wave API: For developers interested in building applications

for Google Wave, be sure to check out the Wave API website

WORKING
Wave offers a user interface suggestive of e-mail and chat but with a
different set of features. A user creates a wave, enters content—a
typed message, an attachment, a web-based widget— and adds other
people to that wave. Although waves appear in a user’s “inbox,” waves,
unlike e-mail, are not sent. Users access waves online, where they can
read and edit existing blips, respond to them, or add new content.
Discussion can be real-time or asyn- chronous and is not necessarily
linear, as those taking part can edit, delete, reply, or insert anywhere in
the conversation. A play- back feature lets participants review the
history of all blips and activities that took place in the wave—like a
movie that shows all the revisions and changes. Wave supports drag-
and-drop of media from the desktop or other applications, such as
interactive maps, trip planners, a weather service, or informal polls.
Adding a blog to the contacts list will allow users to publish a wave to
the blog while retaining complete Wave functionality. Inside a wave,
users can respond to specific blips with “wavelets” (subordinate
conversations) that can be open to all participants or “sent” pri- vately
to selected individuals. Developers are being encouraged to produce
applications that work with Wave through automated processes,
known as robots, and add-ons called gadgets. One Google robot, for
example, allows Wave conversations to occur between people using
different languages with text translated in real time.

Installation and configuration

Stand-alone installation

The Fed One server can be installed by following this manual:

http://code.google.com/p/wave-protocol/wiki/Installation

Several (configuration) choices have been made for the installation of

this server for SURFnet. These are:

 Installed on wave.surfnetlabs.nl, running Ubuntu 8.04.3

 XMPP server is Openfire 3.6.4, installed as mentioned in the
above installation manual
1. OpenFire domain is wave.surfnetlabs.nl
2. Administrator is "admin", password hint: “coffee,
please”
3. Besides the configuration mentioned in the
installation manual, some extra configuration steps
have been performed to the Openfire server, see
"Extra Openfire configuration" section below.
 Wave Extension Installation (this is the actual FedOne server
software which is an XMPP extension):
1. Version from build.properties: fedone.version=0.2
 2. o Installed in /opt/wave-protocol
3. Certificate and key are in /opt/certs, note that you
will also need the CA chain. See the "Extra notes on
configuring the wave server" here below
4. See the complete wave server configuration from
the run- config.sh in Appendix A.

Using the wave server

Starting the wave server

The wave server is started separately from the OpenFire server. Before
the wave server can be started, Openfire needs to run.

The wave server can be started by issuing the following command(s):

$ cd /opt/wave-protocol

$ ./run-server.sh

Note that the default installation does not let the run-server.sh process
detach from the terminal. So, the wave server will terminate when the
user who started run-server.sh logs out. For this POC, the run-server.sh
has been modified so that the java process it starts is detached from
the terminal and all logging is sent to console.log file. See the modified
run-server.sh in Appendix B.

Stopping the wave server means looking for the java process belonging
to the wave server (do a "ps -ef |grep fedone-server") and killing the
corresponding process id. Note that stopping the wave server removes
all waves and users from the server. The wave server does not use
permanent storage for these.
Using the wave client console

The client console can be started on this machine (it will only work from
localhost) with the commands:

$ cd /opt/wave-protocol

$ ./run-client-console.sh <username>

Note that if the username does not yet exist, it will be created
automatically.

Connecting with other wave servers

1) Make sure the wave server is accessible for other wave servers over
the internet. Make sure the XMPP domain and the wave extension of
the XMPP server both have a DNS A record and point to an IP
address on the server. In the setup as described here, we are using
wave.surfnetlabs.nl for the XMPP domain and
wave.wave.surfnetlabs.nl as the DNS name for the wave extension,
both resolve to 192.87.110.93. (note that in a default installation, the
wave extension is called "wave", so this is pre-pended by Openfire to
the XMPP domain name, which in this setup results in wave.wave...).

2) The default port that is used for the server-server communication is

5269 (this can be changed in Openfire). This port should be
reachable for other servers so firewalls and switches/routers should
be configured appropriately
3) For both the XMPP domain and the wave extension, a DNS SRV
record needs to be created:

_xmpp-server._tcp.wave.surfnetlabs.nl. 3200 IN SRV 10 0 5269

wave.surfnetlabs.nl

_xmpp-server._tcp.wave.wave.surfnetlabs.nl. 3200 IN SRV 10 0 5269

wave.surfnetlabs.nl.

Testing with other wave servers

As of November 2nd 2009, the Google Wave Developer Sandbox

instance is open for other wave servers. This means that it now is
possible to connect the installed wave server at wave.surfnetlabs.nl
with the Google Wave Sandbox. Actually, nothing really needs to be
configured for this. Just login to the sandbox at
https://wave.google.com/a/wavesandbox.com, create a new wave and
add a user from the wave.surfnetlabs.nl server (like
[email protected]). The user from the sandbox uses Google's
rich web application whereas the user at the SURFnetlabs wave server
will need to use the console. See the screenshot below that shows the
rich UI and the console with a shared wave

IMPLEMENTING A WAVE ROBOT

Google Wave is positioned as a collaboration tool that (among other
things) enables users to come together and create a document. It
allows users to simultaneously edit the document and let new
participants in the wave use the playback function to see the
document’s history. But when the document is finished inside the
wave, the problem becomes how it could be published to a wider
audience. Usually, a wave does not only contain the document but also
wavelets and blips discussing some aspect of the document while it was
created. The authors might not want to expose these discussions to
others except the wave participants. Furthermore, not everybody has
access to Google Wave or even the wave with the document itself and
not everybody is comfortable with its user interface. This makes it
desirable to, at some point, publish the document to a different
location that can be accessed by anyone in a generally known format
(like plain text or MS Word).

Design

Developers can enhance Google Wave by authoring mini-applications

(extensions) that interact with the Wave. Developers can create robots
which interact with waves, or gadgets which participants may interact
with (and combinations of robots and gadgets are also supported).

For the functionality requested by SURFnet, a robot is the most

suitable wave extension. The main reason for this is that CMIS client
libraries do not exist for JavaScript which is needed when creating a
gadget. There are Java libraries for CMIS available from the Apache
Chemistry project3 and these can be used from a Java robot

 Configure the location of the CMIS repository and the credentials

used to access it
 Configure the document name and folder path
 Mark a specific blip for publishing; as there are potentially many
blips inside a single wave or wavelet, the robot needs to know
which one holds the document to publish
 Publish the document
Implementation

A Java wave robot was created called “CMIS publisher” by using Eclipse
as the development tool. Robots for Google Wave need to be hosted
and for now, only the Google App Engine is supported. The Google
Wave API Java Tutorial4 was followed for creating and setting up the
Eclipse project. The address of the CMIS Publisher is:

[email protected]

This robot consists of several Java source files and uses the Apache
Chemistry CMIS client library for contacting a CMIS repository.

The CMIS Publisher can be used with repositories supporting

version 1.0CD04 of the CMIS standard. Currently, the Alfresco
Community Edition v3.2r2 is the only major CMS to implement this
standard and testing has mostly been performed against this repository

When the CMIS Publisher robot is added to a wave, the root blip is
modified to show the configuration fields for the robot (see Figure 2).
These fields are pre-filled with default values from a publicly available
Alfresco CMIS repository. A wave participant needs to complete these
values before the content of a blip can be published. The text field
below the “Publish document” button is a field that provides feedback
messages for the participants. When not all configuration fields have
been correctly filled or when something goes wrong with the publishing
process, this is where informative, warning or error messages will be
shown.

The user needs to mark a specific blip as the document to be

published. This is done by entering the text “#!publish_me” at the
absolute beginning of the blip. When a participant tries to mark
another blip in the same wave, the CMIS Publisher will remove the just
added tag and warn the user that only one blip can be marked for
publishing.

The content of the marked blip is stored in the App Engine’s datastore
and constantly updated whenever the blip is submitted by clicking the
“Done” button or by focussing the mouse on another wave element

When the participants want to publish the document to the configured

CMIS repository, one of them only needs to click on the “Publish
document” button in the root blip.

1) It will check all values of the configurable fields

2) It will check the datastore to see if this wave has a blip marked for
publishing
3) It will look in the CMIS repository if a document with the same name
already exists
 a) If a document with the same name does not exist, a new
document will be created with the content as stored in the
datastore
b) If a document with the same name already exists, the document is
replaced with the content as stored in the datastore

If everything worked, some informational messages will be shown in

the text field below the “Publish document” button. If the publishing
process fails, an error will be shown.

The CMIS Publisher robot performs extensive logging to the App Engine
logfiles, including the interactions with the CMIS repository. Most of
these

Issues

While creating the CMIS Publisher robot, several issues were

encountered. Some issues were due to the design of the robot, some
due to the limitations of Google Wave and others caused by the CMIS
client or server implementation. Some of these issues impact the
usability of the CMIS Publisher robot and others can be considered to
be notes for a next robot implementation. The implementation of this
robot was not straightforward. This had more to do with the CMIS part
then with Google Wave specifics. CMIS has gone through a lot of
changes in the past months and the available client and server
implementations were specific to a certain version and not
interoperable. This made the search for a correct client – server
combination difficult. Furthermore, the CMIS client libraries (Apache
Chemistry project) were very new and without any documentation and
very few examples. In the very near future, CMIS version 1.0 will be
final and the available implementations will converge to it. This will
make implementing a CMIS client much easier.

As for Google Wave, it is clear that the available API’s do not yet cover
all necessary functionality needed for a robot like the CMIS Publisher.
For this robot, the most pressing issue is the inability to get formatted
text from the blip. Publishing a document to a CMIS repository without
formatting is almost useless. The inability to host a robot on a hosting
platform other then Google App Engine limits the tools available to the
developer. But these are all things Google has acknowledged and is in a
way the result of going public with wave in such an early stage
 Conclusion
Google Wave is not specifically designed to support teaching and
learning activities, but it has the desired features that can be used to
support personalized and directed learning environments. This paper
demonstrates that Google Wave technology can be successfully used to
assists learning according to student’ needs and learning goals of given
engineering courses. Different learning methods during the semester
are utilized, including collaborative learning in groups, working on
students’ projects via sharing, discussing and revising. The Google
Wave platform is flexible and extensible allowing extensive
personalization and customization as needed to tailor the need of
eLearning environment. Its use was well liked and accepted by the
students. A model for personalized learning is created and it leads to
basic understanding of the main factors impacting the personalized
learning, its social aspects, and to students’ assessment of eLearning
technology - what they like, prefers and utilize in practice that
possesses features for learning facilitation in the context of
personalized learning needs. This exploration is worthwhile despite the
decision of Google team to not continue development of Google Wave
further as a standalone product. They are in the process of
incorporating the technology developed for Google Wave in other
Google projects that can find use in PLE. Also, the technology is
available in the form of open source and can be used for free in
different educational contexts

References
http://code.google.com/p/wave-protocol/wiki/Installation

http://reports.jiscemerge.org.uk/Benefits-Realisation/View-
category.html

http://opensourceschools.org.uk/social-networking-elgg-alton-
convent.html

http://teachweb2.blogspot.com/2010/08/personal-vs-personalized-
learning.html

http://www.jisc.ac.uk/media/documents/publications/
heweb20rptv1.pdf

http://hp.telecomtv.com/mobileworld08/collateral/
Service_Delivery_Platform_Mash- up/SDP.pdf

http://www.marklogic-news.com/images/
MarkLogic_Flatirons_07_Using_DITA.pdf