TPG Course Report ================= John Longley, 30 November 2001 Here is a brief report on what has been happening in the TPG course this year, and in particular on my efforts to implement some of the changes that I proposed at the Research Committee meeting in March 2001. Although no specific motion was passed at that meeting, I felt that there was sufficiently broad endorsement of the spirit of my proposals to justify certain changes in that direction. This year's students -------------------- The number of participants on this year's TPG course is unusually large. We have 8 postgraduate students proper, plus 2 long-term visitors, who are engaged more or less full-time on the course. Besides this, a few other people (e.g. a couple of students from Maths) are sitting in on particular courses (although in some cases their attendance has tailed off throughout the term). Many of the students have come with strong backgrounds in theoretical computer science. They all seem committed to the courses, there is a good group ethos, and they generally report that the course so far is demanding, but not excessively so. This year's courses ------------------- The basic objectives of the TPG course have remained as before: * To provide students with a broad background in theoretical computer science and in particular the work done in LFCS; * To prepare students to work on their thesis. To meet these objectives more effectively, I proposed that the course contents should be brought more up-to-date, and in particular that the selection of courses offered should not be regarded as fixed from year to year, but should try to reflect the balance of the research that is actually done in LFCS at any given time. In fact, in order to meet the first objective, it seems one cannot really get away from the traditional selection of first term courses, although this year the contents of some particular courses (e.g. FPLS and, I believe, Concurrency) have been significantly updated. As regards the second term courses, however, this year I am putting together a selection of courses with a greater emphasis on live research topics - hopefully this will feature a larger number of shorter courses. I am also hoping that in view of their timeliness, these courses will also be of interest to other members of LFCS. The Term 1 courses (this term) have been: Category Theory (Daniele Turi) Computation and Logic (Matias Menni) Concurrency (Colin Stirling and Javier Esparza) Functional Programming / Formal Program Development (Don Sannella) Formal Programming Language Semantics (John Longley) The following (traditional) courses will *definitely* run in Term 2 Complexity Theory (Mark Jerrum) Advanced Logic (John Longley) Type Theory and CAFR (Randy Pollack) In addition, Javier Esparza has in principle agreed to give a course on a current research area of interest to him. This would seem a good opportunity for Javier to share his research vision with the rest of LFCS! I am also working on a number of other individuals whom I hope to persuade to give courses (possibly very short, e.g. 6 or even 4 lectures) on current research topics. Supervisor involvement ---------------------- There was broad agreement in March that it would be good to make the course a little more flexible and tailored to the needs of the individual student; also that the student's supervisor should have a little more involvement in planning a course suitable for the student. This year, in Week 0, I held meetings with each student and (in almost all cases) their supervisor, to discuss the course generally, how the student should approach it, and what selection of courses he/she should take. These typically only lasted about 20 minutes and I think they worked well as a simple way of getting the supervisor "on board". In fact, in almost all cases, the recommendation was for the student to do basically all the courses - though in some cases, where students already possessed a strong background, the advice was e.g. "you might not need to go to the first 6 lectures of course X, but you should consult with the course lecturer yourself on this". I have retained the nominal distinction between Compulsory and Optional courses in this year's handbook, but with the qualification that these labels are for general guidance only - in practice, what is "compulsory" for a particular student is ultimately decided by agreement between the student, his supervisor and the TPG course organizer. Timetabling ----------- In accordance with a suggestion received from some previous TPG students, this term I timetabled the courses so that one day (Wednesday) was left free of lectures. The intention was to allow the students to organize their own time more effectively and make it easier for them to find time to pursue their own research interests etc. I haven't yet had any feedback on whether this is proving useful, but it doesn't seem to be doing any harm. Outside courses --------------- We have again made an effort to encourage students to take at least one (usually Informatics M.Sc.) course outside LFCS, either in the first or second term. The choice of course was one of the things discussed in the meeting with the supervisor at the beginning of term. However, in practice we often found ourselves beset with timetable clashes, and it is not clear what should be done about this. Possibly I could have spent more time looking at the M.Sc. timetables and thinking about which courses people were likely to want to go to; but I think my efforts were hampered by the fact that other course organizers tended to leave things just as much to the last minute as I did. I do not know how many of the students have so far actually been going to outside courses.