Final paper due on Dec. 4

After our discussion on Wednesday about reviews, I thought you might be interested in the process. What we will do is the following:

• On Dec. 4 you submit 6 copies of your paper (hardcopies), bring them to class. Dec. 4 is a hard hard deadline.
• I will assign each paper a number.
• We divide them across the class. Each of you will read 4 papers. (Make sure that you identify conflicts: ie. you are not allowed to review your own paper or a paper of your friend.)
• After reading them, you will fill out a review form for each paper assigned to you, and send them to me by email by Dec. 9, midnight.
• I'll rank the papers on their collective scores, tabulate the results and reports, and we discuss the top papers in class.

This process is pretty similar to the process that a real _systems_ conference would follow. (The procedure differs from field to field.) In fact, the enclosed review form is the one that has been used for OSDI, ASPLOS, SOSP, etc.

The area for specific comments is the space where you can identify specific strenghts and weaknesses of the paper. The typical structure of this section is a one or two paragraph summary of the paper (so that the rest of the committee can figure out what the paper is about), followed by your review.

Some people write long reviews, some write short ones. The comments are intended to be constructive; they go back to the authors so that they can improve their papers. So comments like "The idea in this paper is stupid; I do not like it" are bad comments; comments of the following form are fine:

The ideas in this paper are fantastic, but the implementation does not support the stated claims. For example, in Section 7 the authors state that the world is square but from previous experiments (see the paper "The world is a triangle" by Smith in the last SOCS) we know that the world is a triangle. I therefore think we should reject this paper.

Please, if you have any comments about the process and review forms, send me email or raise them in class.

I strongly encourage everyone to submit their project paper, even if your project is not completely finished yet, or your paper is not as strong as you wish it would be. First, idea papers can be very interesting (e.g., active networks paper). Second, you will not be the only one who has an unfinished project. Third, the reviews might contain useful comments that might help you in strenghtening your work (e.g., for your thesis). Four, submitting is required if you want a passing grade for 6.853.

Reading instructions

More detailed reading instructions for the next couple weeks are available at the reading list page

Handouts

The following handouts are new or current. Choosing "" will fetch a PostScript version of the handout. Most handouts are also available in HTML, by choosing the name of the handout itself. Hardcopy-only handouts can be obtained from the course secretary.

  1: 6.853 at a glance (frame)