Internet2 "Neternship"

Summer, 2005

The focus of this project is to provide students with the opportunity to collaboratively create or enhance a piece of software that takes advantage of the advanced network features provided by Internet2.

Goals

• Provide a small group of students the opportunity to interact with students and faculty from other campuses through a distributed programming project
• Introduce students to collaborative tools that are currently in use across Internet2 and the commodity Internet
• Produce functioning software by the end of the term that addresses all functionality identified at the outset of the project

• Hands-on experience for students in advanced network programming fundamentals
• Collaborative experience will put students in touch with new communities: network infrastructure, video, collaborative tool providers, researchers, etc.
• Introduction to Internet2 (technology, staff and community) will be beneficial not only for the participating students, but also for campus staff. Contact points deep into the Internet2 community will be useful for future questions regarding networking at the campus and backbone level.
• Opportunity not only for students, but for campuses to to contribute to the advancement of networking as a whole. The tools and technologies created will be publicize by Internet2 and can be cross-referenced for local announcements.

• The students will participate via independent study on their own campuses. Final grading and credit will be determined by each student's local campus instructor/advisor. Students are expected to be self-motivating and know when to ask for help.
• Students will participate in a weekly "class" using the Internet2 Commons video conferencing infrastructure (H.323), AccessGrid, VRVS, and/or ConfXP.
• Students are expected to be proficient in network programming and familiar with compile/debug tools for the languages. Languages possibilities will probably be C, C++, and/or Java, but specifics will be determined once the student group is assembled.
• Each student's instructor will be available for one online "class" instruction. The instructor will provide Q&A, guidance, and consultant-style assistance.
• Software will be produced under the Internet2 Intellectual Property Framework
• Software compilation should rely only on open source software. Specifically, open source tools should be used for compilation, linked libraries, etc. Software can be created using any tools (e.g., closed source editors).

Interesting and Useful Additional Skills?

• Database knowledge
• Network routing
• Graphics programing (UI and 2D, do not expect 3D will be useful)
• Glue languages like Perl, Python and/or Ruby might be useful.

Potential Project Ideas

Internet2 Detective

• Currently, there are at least 3 source trees (in at least that many languages). It would be useful to consolidate these into one main trunk. The core code is envisioned to be C (via gcc) and the cross-platform GUI could be implemented through wxWindows. This would also be a good opportunity to look at the ways in which the code could be modularized/layered to more easily support expanded functionality.
• The Internet2 Detective currently covers a small number of network tests to provide insight into a host's bandwidth capabilities. Adding diagnostic features (on a per-platform basis) would move this from a measurement to a diagnostic tool. Initial work that can be leveraged is available in Java from the NDT, Network Diagnostic Tester.

DigitalVideo (DV) and the AccessGrid (AG)

• AG is a multi-site, multi-camera, video-conferencing tool that has traditionally relied on H.323. Improvements and cost reduction in consumer hardware now provide an opportunity for higher quality video. DV support is available for Linux. The Windows and OS/X implementations are sub-optimal, but (barely) functional source code exists. Optimizing the DV tree would provide existing users with significant improvements in quality while opening the door to an entirely new class of users.

Collaborative Tool

• A variety of collaboration tools exist and most do not integrate well with one another. VRVS's reliance on VNC is one of the more plausible, successful examples. It would be interesting to provide a cross platform mechanism to collaborate on documents such as the Mac specific tool subEthaEdit. (TCY has specific ideas on this, but has many possibilities and may not provide enough constraints for a good project.)

Calendar

Oct 2004

• Initial distribution of this proposition
• Identify instructors who are interested in participating
• Through instructors, identify potential students

Nov 2004

• Select 2-4 potential project ideas
• Through instructors, recruit and select students

Dec 2004

• Identify instructors and students who will participate
• Create calendar for participants (start dates, meeting times, end date, demo opportunity, etc.)

Jan 2005-May 2005

• "Neternship" in progress
• "Neternship concludes with final presentation