During the course of the research

Throughout the course of the research, keep in mind the objectives of the thesis, and the instructions in "A Simple Approach to Thesis Writing". A thesis is not like an assignment: it has no visible end, and it is acceptable to change the problem, if this allows you to make a more interesting claim in the end. A change would be incremental, not radical, and it should take advantage of your investment in literature research and software that you have already made.
It is important to keep your goal clear in mind, and make your research the highest priority in your life at this time. Progress is sometimes fast, sometimes slow, and sometimes non-existent. Make an effort to have a minimum of progress towards your thesis every day. Meet with your supervisor often. Weekly meetings are strongly recommended when your thesis is defined (or redefined), and when it is being wrapped up and the claims are being fine tuned. In between, less frequent meetings are acceptable.
Do not disappear for weeks on end with no meetings, and with no concrete progress being made towards agreed upon goals.

Here is a proposed system for keeping in touch and keeping track of progress:

For each of my students, I have a directory: locutus:~eem/research/<yourLoginID> also accessible over the
Web as www.cs.dal.ca/~eem/Research/<yourLoginID>
This directory has been included in the global robots.txt file, so it should not be indexed by search engines.

In that directory, I keep:
- drafts of reports, theses, papers. This is the default location where I upload the pdf files annotated with my comments for you.
- a plain text file Schedule.txt, that contains the current best estimate of milestones towards completion of thesis
- a directory Readings, that contains research papers relevant to your research.

How long does it take to generate the first draft of the thesis?

This depends on how good your bookkeeping has been during the course of the research. The following hints will speed up the process:

• Write up algorithms and experimental results as soon as they are completed. This way when thesis writing time comes, you already have important fragments of text for the thesis, instead of having to write all the text from scratch.
• Do a presentation of the completed research before starting to write the thesis. More advice on this in step 1 here.
• Read, understand and internalize the instructions in "A Simple Approach to Thesis Writing". If you have any questions, discuss them with me before you start writing.

Your first draft should be complete. A complete draft means that it includes the full text of the thesis, with no parts missing, and the experimental results have been thoroughly discussed and put in their proper context, following the instructions in "A Simple Approach to Thesis Writing".

Your thesis should be written in LaTeX. In addition to other students in the Faculty using LaTeX, here are Resources on thesis typesetting in LaTeX. LaTeX may have a higher startup cost compared to MSW, but it trivializes the typesetting effort of revisions, and generation of publications for conferences and journals. If there is a critical number of students interested, I can give a tutorial presentation on LaTeX.

The first draft of the thesis you submit to me should be your best effort, and you should review it critically yourself first.
Please keep in mind that writing the thesis is not a collaborative exercise with your supervisor. You are fully responsible for the writing. Your supervisor is there to guide you through the process and give you feedback. The draft reading capacity of your supervisor is about three drafts. Do not expect your supervisor to help you turn rough notes into a thesis through an infinite number of iterations on bits and pieces of the thesis.

How long does it take to turn the first draft into a completed thesis?

Experience has shown the average time required to fine tune the first complete draft of a thesis into a defensible version is two months. During this time, several iterations on the text of the thesis will take place, plus additional experiments will be conducted to strengthen the arguments in the thesis.
A thesis defense will only be scheduled after I have reviewed and approved a complete draft of the thesis.
The defensible draft results from a good first draft after 2-3 revisions.

Please take this into account when you plan your graduation or accept a full-time job, and do yourself a favour by allowing for the time required to get a decent thesis done.

How do I receive and respond to comments on my thesis (or paper) draft?

To make most efficient use of your and my time, we will follow a "protocol" for interacting on the changes. I read and comment on thesis drafts electronically on pdf files using Acrobat. The protocol is described in the following link.
How to get from completed experiments to a completed thesis.