Project Information

CMSC 621
Spring 1997

Overview

The project for this class will be a semester-long effort involving 2-4 students working together in a group. Each person in the group should expect to spend 50-70 hours over the course of the semester working on various aspects of the project. All projects will involve additional background research beyond the paper readings for the class, and all groups will be required both to present their work to the class and submit an 8-12 page paper on their project.

Project topics may be chosen from any area of operating systems, but they must involve some original work by students. This doesn't mean that each group must come up with a new idea ­ groups are welcome to verify previous research or compare different approaches to a problem. However, such comparisons should involve quantitative rather than qualitative studies, since merely summarizing several papers in operating systems isn't a project. Students are encouraged to do projects that explore new ideas in operating systems ­ the further the idea is from current work, the less quantitative work is necessary to support the paper. This is largely done because of the limitations of time and because of the difficulty of implementing radically different ideas in a few weeks. The instructor has final approval of project topics, but every effort will be made to make sure that students are working on projects that interest them.

Checkpoints

To insure that groups are making steady progress towards their goal, there will be regular checkpoints during the course of the semester. Checkpoints should not require a lot of work beyond normal progress on the project; rather, they are intended to keep groups on schedule to complete the project by the end of the semester.

Checkpoint 1 (due 3 March 1997):

At this checkpoint, each group should hand in a tentative project title with a short description. Each group will meet with the instructor for 10 minutes to discuss the project topic; after the meeting, the group should be able to begin doing the background research for their project.

Checkpoint 2 (due 19 March 1997):

By this time, groups should have done the background research for their project. This can include papers from a number of sources including journals, conference proceedings, and technical reports. The WWW is a particularly good place to look for articles, but it should not be the only place to look. For the checkpoint, each group should hand in a list of papers (in a standard bibliography format ­ see one of the papers for an example) relevant to their project. There should be at least 8-10 papers on the list; more papers are of course welcome.

Additionally, groups are welcome to include any preliminary progress on the project design in this checkpoint. While such progress isn't required, it will make it easier to finish the project by mid-May. This may also be a good point to get feedback from the professor on the preliminary design.

Checkpoint 3 (due 16 April 1997):

This

Presentation

(In class, May 7-12, 1997)

Presentation info goes here.

Paper

(Due in class May 12, 1997)

Paper info goes here.