This document provides an overview of open source technologies that can be used in science education. It discusses definitions of free and open source software, examples of open source operating systems and applications, and specific tools for tasks like numerical computing, data analysis and visualization. The document also shares the author's experiences using open source technologies like Gentoo Linux, and concludes that open source is well-suited for education due to benefits like low costs, reliability and allowing students to examine internal workings.

1. Open Source Technologies in
Science Education: What's Your
Geek IQ?
Stephen L Arnold
Assoc. Faculty, Allan Hancock College
Gentoo Linux Developer
Sci, dev-tools, comm-fax herds
stephen.arnold@acm.org nerdboy@gentoo.org

2. Overview
• Definitions
• What is Free and Open Source Software?
• Overview of the technology
• Examples
• My View from the Trenches
• Specific applications
• Student scores
• Results and Conclusions
• What works and what doesn't
• Open source to the rescue
• Gentoo Linux (extra)

3. Why Open Source?
• User Freedom - Examine, use, and extend to fit
your needs.
• Life-cycle costs - Reliability, Availability,
Maintainability (RMA)
• Pedagogy - Internals open to all, supports
multiple levels of instruction, verifiability.
Citing everything from insufficient education and training,
to the perceived but unrealized benefits of technology, and
concerns about the risks to society, the NRC reports a need
to get more value from information technology and
education (NRC, 1999). NRC report also directly links
wages to technology skills and education.

4. Open Source Technologies
• Earth Sciences
• Meteorological Analysis & Modeling
• Image Processing, GIS, & Mapping
• Data Formats & Transports
• NetCDF, HDF4/5, LDM
• General Scientific Software
• Numerical Computing/Libraries
• Cluster/High Performance Computing
• Analysis/Visualization
• Desktop Tools
• Document Production – SGML, XML, PDF
• Graphics Manipulation – too many to list
• Archiving & backup – tar, gzip, bzip2, zip

5. Some Real-World Examples:
RedHat with Gnome

6. Gentoo Linux Gnome
Desktop

7. Open Source Technologies
(cont.)
• Operating Systems and Hardware Platforms
• Linux – Numerous distributions that run on x86, PPC,
HPPA, MIPS, ARM, Sparc, Alpha, s390
• BSD – OpenBSD/NetBSD/FreeBSD (at least x86)
• Various Embedded Devices – Sharp Zaurus, Ipaq, etc
• Network Services
• Web Application Servers, other services
• Internet clients – web, email, messaging
• Security tools and services

8. Software Development
• Programming Languages
• Ada, Python, C, Fortran, C++, Java, Perl, Ruby, AWK,
SmallTalk, Lisp (and more)
• Development Tools & Libraries
• GCC, SourceNavigator, Eclipse, Doxygen, Emacs,
bugzilla
• Glade/Python, GTK/CORBA (Gnome), Qt (KDE)
• Udev/sysfs, HAL
• Configuration Management
• CVS/RCS, subversion, arch

9. View from the Trenches
• Gnome Desktop
• provides integrated yet independent components and
applications (CORBA, XML)
• Nautilus file manager for local and remote resources
– FTP, SSH, SMB, WebDAV
• OpenOffice productivity suite – Open document
formats, import/export
• Mozilla/Firefox – Browsing, email, chat, XML
tools
• Accessories, games, multimedia, system tools
• GIMP – GNU Image Manipulation Program

10. Gentoo with Gnome 2.8.2

11. Network/Web Services
• Apache/Zope Web servers:
• Documents, Applets, Discussion Forums; Future
Directions: Content Management and Portal services
• Real-time Chat and Conferencing:
• IRC as Virtual Office Hours, Video conferencing with
GnomeMeeting
• Third-Party Web Services:
• Blackboard, textbook companion sites, other
academic web sites
• Gentoo Linux:
• Supports scientific & high performance computing,
as well as general education, with a huge collection
of cutting-edge applications

12. In-house Web Hosting
• Scalable and flexible
• Author's geography courses currently hosted on
personal web server on home DSL connection
(dedicated geography server in purchase order
queue; total cost $600 plus build time)
• Additional planned services include live
meteorological data, web-based analysis tools, as
well as web-based GIS & mapping tools (requires
additional network infrastructure and bandwidth)
• Low Cost
• Hardware and bandwidth only
• OS and network servers freely available
• Test on the desktop, deploy on the LAN
• Some Zope Examples

13. Zope Web Application Server

14. Zope Management Interface

15. Editing a DTML Page

16. SquishDot: An Example
Product

17. Managing SquishDot

18. Zope CMF: The Content
Management Framework

19. Assessment Types
• Exams, group/individual assignments - on-line
applications, some traditional
• Required: Open-book exams, group and individual
assignments and activities, semester project (eg,
contributing relevant articles and comments on a
web-based discussion forum)
• Optional: Additional articles, web-based quizzes and
exercises provided by textbook publisher
• Successful assessment methods combine
peer-interaction and community
• Discussion forum project
• Migration research project
• Focused group discussion (non-virtual)
• Simply using the WWW is not enough

20. Assessment Results
Student Geography Scores by Semester
Semester LOW HIGH MEAN MEDIAN STD-DEV N
Fall 2001 45 84 70 76 13 16
Fall 2002 50 87 68 67 12 14
Spring 2002** 35 95 82 87 16 12
Spring 2003*,*** 35 97 84 89 16 23
Fall 2003* 82 94 90 91 7 9
Spring 2004* 61 100 90 92 17 26
Fall 2004 21 99 76 84 32 17
* includes web-site extra credit
** first use of Squishdot discussion forum project
*** migration research project

21. Training and Logistics
Issues
• Student Technology Proficiency
• Medium on average, however, with a very broad
spread (ie, from very low to very high). Universal
technology courses recent and not wide-spread
• Basic Skills
• Also highly variable in all areas (student population
high in recent immigrants and other ESL students)
• More Training Required
• Schools beginning to require freshman technology
courses
• Faculty/Staff professional development

22. Technological Conclusions
• Excellent overall
• Low cost, reliable, stable, cutting-edge tech
• Everything you've seen today was created with
GNU/Linux and other Open Source tools
• Philosophically compatible with public education and
research
• Open and verifiable
• Highly secure and flexible
• Puts the user back in control
• Ideal for Students and Educators
• Freely available and distributable
• Runs on older hardware of many varieties
• Compatible with other platforms

23. Pedagogical Conclusions
• Standard assessment tools not very good
• Seductive and seemingly easy, but very little to
engage students
• Several semesters of low scores speak for
themselves
• Open Source tools foster creativity
• Customized and targeted web applications in just a
few mouse clicks
• Allows for both automation and exploration
• Appropriate technology use is key
• Importance of technology and environment
• Real-world vs. virtual
• Connecting course topics to the customer
• Peer Interaction and “a place of their own”

24. Fini
Thanks for coming, and thanks for bringing some
weather to San Diego... :)
Further reading:
http://www.gentoo.org/doc/en/index.xml
http://www.python.org/moin/BeginnersGuide
http://www.python.org/sigs/edu-sig/
http://www.stgray.com/quotes/python.html
http://www.zope.org/
http://www.zope.org/Documentation/Books/ZopeBook/

25. Gentoo Details
• (extra slides)

26. Gentoo Linux Features
• glibc 2.3.4+, gcc 3.4.x, ext3, ReiserFS, XFS,
JFS, ALSA, pcmcia-cs, SELinux,
PaX/Grsecurity, Propolice
• Xfree86 4.3 (and xorg 6.8.0-r4), OpenGL, KDE
3.3.1, Gnome 2.8, iptables, QoS
• qmail (with mysql and LDAP), postfix, courier
• udev/hotplug, sysfs, GRUB, lilo, milo, silo, palo,
yaboot, BootX
• Multiple kernel possibilities
• Vanilla 2.4.26, 2.6.8.1 (stable), aa-, ac-, acpi-, mm-,
openmosix-, ppc-, ppc-benh-, redhat-sources
• Gentoo kernels with preempt, low latency, EVMS
• Prelinking, ccache, distcc

27. Gentoo Portage Features
• Provides functions/scripts which download,
patch, compile, and install packages
• Modeled on the ports-based BSD distributions
• Dependency checking, extreme customization
• Original source tarballs are downloaded
• The user specifies what they want, and the
system is built to their specifications
• Compiles are optimized for your specific hardware
• E.g. Altivec on G4 PPC chips
• Pentium versus Athlon
• Specify settings once, and all packages are built to
those options

28. Portage Ebuild Scripts
• Easy to read format, clear separation of
phases
• KEYWORDS, DEPEND
• Stable versus testing
• CPAN/Portage integration
• Install and manage dependencies via Portage
• Package management for perl modules
(including uninstall)
• Ebuilds are automatically created for CPAN
packages

29. Portage USE Flags
• Globally defined list of features
• Configure yours in /etc/make.conf
• USE="-gnome kde qt arts -nls python perl
oggvorbis opengl sdl -postgres jpeg png
truetype dvd avi aalib mpeg encode fbcon
mmx"
• Each one defines specific functionality for
each package to support
• USE flags generally map onto --configure
options
• Install only what you want. No need to trim
down a default installation
• “Opt-in” versus “opt-out”

30. Installation Process
• Currently no graphical installer
• Just follow the detailed install documents
• Briefly:
• Boot from CD, setup networking, partition
• Unpack stage 1, stage 2, or stage 3
• Chroot, bootstrap or emerge system
• Compile kernel, install system logger, cron
daemon
• Setup bootloader.
• Set timezone, configure additional users

31. Stage Tarballs
• Stage 1 install
• Bare-bones. Need to bootstrap, compile gcc, glibc,
system (make, perl, etc.), kernel, and user
environment
• Stage 2 install
• Already bootstrapped. Compile system, kernel,
and user environment
• Stage 3 install
• Base system included. Compile kernel and boot
manager
• GRP install (Stage 3 add-ons)
• Precompiled packages with the default MAKEOPTS
and USE flags for your architecture

32. Binary Packages
• Build your own, distribute packages to your
machines
• emerge --buildpkg
• FEATURES="buildpkg"
• PORTAGE_BINHOST=”http://local-server”
• GRP (Gentoo Reference Platform)
• Pre-built binary packages using default options
• Including Xfree86, Mozilla, Gnome, KDE, Emacs,
OpenOffice.org, Apache, MySQL, PostgreSQL,
Samba

33. Try Gentoo Linux
• Download from www.gentoo.org
• 95-600mb iso images, plus 10-85mb stage files
• Unreal Tournament 2004 demo for x86/NVIDIA
• Does not touch your hard drive
• Live CDs for x86, PPC (NewWorld and
OldWorld), Alpha, AMD 64, HPPA, Sparc
• All live CDs are also install and rescue CDs

34. Documentation
• Gentoo Handbook
• Installation, FAQs
• Portage user manual
• USE flags, ENV.D, Security guide
• Desktop configuration guide, rc-scripts,
ALSA, DRI, Java
• AFS, OpenMosix, Diskless/LTSP, Printing,
UML, IPv6, Virtual Mailhost
• Developer documentation
• Ebuild creation, eclass
• Documentation guide (XML syntax)

35. Gentoo Community
• Close contact with end users
• Many ebuild scripts are submitted by users
• IRC channels (on irc.freenode.net) - #gentoo
is the largest on the network with 900+ users
• Web-based forums (on forums.gentoo.org),
1000+ posts per day, 100,000+ topics
• Fully-public bug tracking (bugs.gentoo.org),
20,000+ hits per day
• “Gentoo Bug Day”
• Linux World Expo in Boston, SoCal Linux
Expo in Los Angeles (Feb 2005)

36. Staying Informed
• News on www.gentoo.org
• Gentoo Linux Security Announcements
(GLSA)
• RDF feed of news. GLSA feed coming soon
• Multiple mailing lists (each architecture,
documentation, security)
• Gentoo Weekly News (GWN)
• Informal Discussion and Announcements
• Gentoo forums on forums.gentoo.org
• Gentoo IRC channels on irc.freenode.net