The Success of Open Source

1
The Success of Open Source
And the Rise of Community Source

• Brad Wheeler
• Indiana University
• bwheeler_at_indiana.edu

2
?
3
Questions for Higher Ed

• Can I trust open source applications today?

• Can I trust open source applications tomorrow?

• Who will support it?

• Should I invest or be a free rider?

• When?

• Who will I hold accountable (sue)?

• Are there really any total cost of ownership
  advantages?

• Isnt vended software safer?

• Does this fit my institution?

Code Coordination Community
4
Success?
All computing/research images from Indiana
University sites
5
The Code
Code Coordination Community

• Linux and Apache have delivered the code can
  higher ed deliver the application code?

6
The Story of Two Codes

• Oncourse code
• Dim targetName
• CourseList ""
• If UCase("" user) UCase(""
  Request.Cookies("OC_COURSE")("DPT")) Or UCase(""
  user) "OCADMIN" Then Exit Function
• Dim mail_dict, mail_ids, read_ids
• Set mail_dict CreateObject("Scripting.Dictionar
  y")
• mail_dict.CompareMode 1 'Text Mode
• conn.open cStr(Application("oc_Courses_conn"))
• Sakai code
• public void init(ServletConfig config) throws
  ServletException
• super.init(config)
• m_redirectContext config.getInitParameter("red
  irect-context")
• m_redirectPath config.getInitParameter("redire
  ct-path")
• m_logEnabled "true".equals(config.getInitParam
  eter("log"))

7
Sakai software in production use at Michigan
Indiana
Does this code meet the current and future needs
for faculty and students?
8
Concurrent Development Leverage Commits To CVS
9
From The EDUCAUSE CIO Constituent Group Listserv
On Behalf of .. Sent Thursday, November 18,
2004 457 PM
Dear Mr. Ellison, we are not interested in
migrating to Collaboration Suite and find
ourselves locked out of an upgrade path for our
Calendar services. If Oracle is unwilling to
sell and support Calendar as a standalone service
we will be forced to migrate to one of the
competing calendaring systems. This would be
unfortunate both because Oracle Calendar is a
quality product that has served us well, and
because of what it says about Oracle's
willingness to listen to and accommodate the
needs of its higher education customers. We are
formally requesting that Oracle commit to
maintaining the standalone version of Oracle
Calendar in rough parity with the collaboration
suite calendar component. We would appreciate a
response to this request by January 10, 2005,
even if it is not technically feasible to deliver
the updated stand-alone version in that
timeframe. Sincerely, ltlist of names,
institutionsgt
10
Control of Code Destiny
But with responsibility too!

• Functionality of Systems
• integration, standardsinnovation
• Cost of Systems
• operations, maintenance, timing, evolution

PKI Dartmouth
11
From Tai, Raymond Y. mailtorytai_at_ucdavis.edu
Sent Monday, November 29, 2004 837 PMTo
sepp-develSubject Sakai beyond
Oracle/MySQL Hi, As I am converting Sakai to use
MS SQL/Server as its backend, I realize this may
be a futile effort on my part unless we have some
direction from the developers/board members of
Sakai as well as uPortal. The following two
questions came to mind (a) Currently, uPortal
supports the following db backends
Oracle, MySQL, HypersonicSQL, PostgreSQL, DB2 and
Sybase SQL Server at least up to the current
version of 2.4.1. If uPortal 3.0 is part of the
Sakai project, then what is the direction of db
backend support? Scale back uPortal db support
to match those with Sakai (Oracle/MySQL) or
expand db support for Sakai to include those
currently supported by uPortal? (B) Currently,
there are db specific code in some of the java
code. Question is how to take these db specific
codes out of the java code if the direction is
to expand db support for Sakai? Any insight is
great appreciated. RT
12
Code Evolution?
Questions of Coordination Questions of Community

• Quality?
• Feature development?
• Cost (TCO)?

Release 2.0
Release 1.0
1.1 1.2
Time

13
The Coordination
Code Coordination Community

• Linux and Apache have demonstratd successful
  models of coordination can higher ed?

14
(No Transcript)
15
Higher Ed Applications
Higher Ed Applications Linux/Apache
Scope Scale Niche market, small vertical Unbounded across industries, nations, large horizontal
Requirements Specialized Generic
Resources ?? 1,000s high quality volunteers
Passion ?? A.B.M.
16
Code Coordination Models
Stakeholder Coordination Open IP
Community Source Projects
Bundled IP Support
Partnering Organizations
17
Consolidation
Commercial
18
Community Source Projects
Community source describes a model for the
purposeful coordinating of work in a community.
It is based on many of the principles of open
source development efforts, but community source
efforts rely more explicitly on defined roles,
responsibilities, and funded commitments by
community members than some open source
development models.
Institutional Investments for Institutional
Outcomes
www.sakaiproject.org
19
Project (product?) Governance?
Governance Specifying the decision rights and
accountability framework to encourage desirable
behavior to achieve an objective
Adapted from IT Governance, Weill Ross, 2004
20
(No Transcript)
21
(No Transcript)
22
Higher Ed Coordination

• Can higher ed capture economies of scale in
  software creation and maintenance? The 100
  Million question

Lifecycle System Costs/ Effectiveness

• Capturing Industry Leverage
• Learning how to partner
• Synchronizing institutional
• investments
• IT architecture discipline
• Creating effective communities

?
Number Participating
23
From Charles Severance mailtocsev_at_umich.edu
Sent Tuesday, November 23, 2004 417 PMTo
sepp-develSubject Some Development Needed
News/RSS I have a request for some development
help. The core team is furiously working on the
1.5 release (Rob Lowden's Holiday gift) and we
don't have time to work on a tool which in my
mind needs some help. The "news" tool is a
legacy tool that does RSS feeds. In my testing
and talking to the original developer of the
code, it is a pretty simple/basic implementation.
I figure that there might be someone out there
who could test it out a bit, and teach it a few
more formats than 0.91 RSS. This is not a
critical thing, but someone with some RSS skills
might find it pretty easy to do. If you are
interested - let me know - so we don't have 2
people in the same place. Here are the areas in
CVS that you would work in. ./chef-tool/src/java/
org/sakaiproject/tool/news ./chef-tool/src/reg/che
f.news.xml ./chef-tool/src/webapp/vm/news ./chef-t
ool/src/webapp/vm/news/chef_news-Control.vm ./chef
-tool/src/webapp/vm/news/chef_news-customize.vm ./
chef-tool/src/webapp/vm/news/chef_news-Layout.vm .
/chef-tool/src/webapp/vm/news/chef_news-List.vm /C
huck
24
From Chris Brandt mailtocmbrandt_at_ucdavis.edu
Sent Wednesday, November 24, 2004 219 PMTo
sepp-develSubject News / RSS reader I mucked
around a little bit in the code and found that
the reason that most rss feeds are being denied
in Sakai is that very few of them contain an
image node (which is required by the current
codebase). By commenting out lines 444-447 (of
legacy-component\src\java\org\sakaiproject\compone
nt\ legacy\news\BasicNewsService.java) and only
setting the m_imageX properties if the image node
actually exists, most RSS and RDF feeds _are_
usable. I've checked several sites including
slashdot and they seem to work just
fine. Enjoy! Chris /if (imageList.getLength()
! 1) throw new NewsFormatException("File
received from " source " is not in RSS
format") / if (imageList.getLength() 1)
image imageList.item(0) ---- snip
---- m_imageLink getNodeValue(image, "link",
null) m_imageDescription Validator.stripAllNew
lines(m_imageDescription)
25
-----Original Message-----From Charles
Severance mailtocsev_at_umich.edu Sent Tuesday,
November 30, 2004 314 PMTo sepp-develSubject
Re News / RSS reader Chris This is awesome!
Thanks a bunch. Your fix is right on the money.
I just tested it and checked it into CVS. I
attach the CVS commit message at the end of the
message. Thanks a lot. Now I need to find more
tasks to do. /Chuck
26
The Collaborative Capability
Organizational capabilities are bundles of
reliable and replicable routines that achieve a
valuable purpose
No not invented here
Timely signaling of local matters
Collaborative Capability
Greater good over just the local good
Distributed work via distance
Earning trust
27
IU Community Source Projects
2004
28
The Community
Code Coordination Community

• Linux and Apache have cultivated worldwide,
  sustaining communities can higher ed?

29
Licensing Frames Community

• GPL/GNU Viral
• Code
• Open source
• Use without fee
• Derivative works
• MUST have same license rights
• Open source
• Use without fee
• Allow derivative works

• Open-Open
• Code
• Open source
• Use without fee
• Derivative Works
• Any license..

Birkenstocks and Wingtips Open Source
Licenses, EDUCAUSE REVIEW, Jan/Feb 2005.
30
Licensing
Wingtips Capitalist, Private good philosophy
Birkenstocks Free software, Public good
philosophy
31
From Fenlason, Clay D mailtoclayf_at_bu.edu
Sent Friday, September 24, 2004 914 AMTo
advocacySubject Small School Sakai Yesterday's
NERCOMP gathering led by Babi, Mark, Jeff and Jim
reminded me of what appears to be a conventional
point of view, but one I've tended to be puzzled
by. Stripped to its essentials, it goes
something like "Sakai implementation is fine
for big schools with lots of resources, but the
cost of entry (esp. in terms of staff) is just
too high for small schools. This sort of thing
always leaves me with a wrinkled brow, since we
became a SEPP partner precisely because we're a
small school with limited resources. A larger
school might be able to operate with fair success
under the illusion that they could go it alone,
but we can afford no such pretension. Further
yet, we don't yet have the knowledge or skills
in-house to just pick up and run with the public
code releases. We need a little more guidance,
training, time to get ourselves up to speed and
forge some collaborative relationships. My first
instinct, then, is to think that those who bemoan
the frailty of the small school are missing the
"community" part of the message they're thinking
"open source," but not "community source."
32
Communities
Core Governance
33
53 Sakai Partners as 31 Oct 04

• Arizona State
• Boston University School of Management
• Brown University
• Cambridge University
• Carlton College
• Carnegie Foundation for the Advancement of
  Teaching's Knowledge Media Laboratory
• Carnegie Mellon University
• Coast Community College District (Coastline
  Community College)
• Columbia University
• Community College of Southern Nevada
• Cornell University
• Dartmouth College
• Foothill-De Anza Community College District
• Harvard University
• Johns Hopkins University
• Maricopa County Community College District
• Nagoya University
• New York University
• Northeastern University

• State University of New York
• Tufts University
• University of Amsterdam
• University of Arizona
• University of California Berkeley
• University of California, Davis
• University of California, Los Angeles
• University of California, Merced
• University of California, Santa Barbara
• University of Cape Town, South Africa
• University of Colorado at Boulder
• University of Delaware
• University of Hawaii
• University of Illinois
• University of Lleida
• University of Melborne
• University of Minnesota
• University of Oklahoma
• University of Toronto

1-2 per week since 1 March -- Others are in in
various stages of contracting
34
The Success of Open Source
Code Coordination Community
And the Rise of Community Source

• Brad Wheeler
• Indiana University
• bwheeler_at_indiana.edu