Minutes: IOI Scientific Committee Meeting 2 Status: Approved (October 2000) Date/place: July 21, 2000, Beijing, China Present: Ismail Hakki Toroslu Host n-1 Wenhu Wu Host n Baolin Yin Host n Tom Verhoeff Elected (chair) Gyula Horvath Elected Absent (with notification): Jyrki Nummenmaa Host n+1 Mohammad Ghodsi Elected 1. Opening The meeting was opened by Tom. Mohammed did not attend the review week because of personal circumstances. It was agreed that Mohammad would arrive early at IOI2000 to provide a fresh opinion on the final versions of the material. Jyrki did participate in the review process, but he did not feel well enough to attend this ISC meeting. 2. IOI2000 Competition Material The preceding days (starting July 17), the ISC has worked together with the IOI2000 SC to review all the IOI2000 Competition Material. A written report about the findings of this review was presented to the IOI2000 SC. The report consisted of a general part and a task-specific part. Each of the recommendations was briefly discussed and, where necessary, explained further. This report is confidential. A summary of the review week will be sent to the IOI IC. The work of the ISC was not finished during the review week. In particular, Jurki would continue to work on improving the English formulation of the tasks; Gyula and Tom would continue to work on the reactive tasks (libraries and test data). Tom, Gyula, and Mohammad will arrive early at IOI2000 and will provide last-minute feedback. 3. Evaluation of the review We evaluated the past review week. Before the actual review was conducted, we did not know whether it would be useful, what would be the best location, when would be the best moment, and how much time would be needed. At the end of this review week, it was obvious that the review was very successful and helpful. It is also clear that the review must take place at or very close to the location of the IOI it concerns. There are several reasons: * The review process requires close cooperation, both among ISC members and with the Host SC members and its technical staff. This cannot be arranged easily and quickly when the persons involved do not meet face-to-face. * The review process requires intense concentration on the job at hand. Traveling away from one's normal duties helps reduce distractions. * A visit to the competition site and demonstrations involving (some of) the actual competition equipment and software are an important part of the review. * During the upcoming IOI, the ISC members may have to carry out some further work. Familiarity with the environment and staff ensures that they can work efficiently without wasting time. Concerning the moment and duration there was somewhat less concensus. Hakki suggested that an earlier (4 months before the IOI) and shorter (3 days) review meeting might be enough and should mostly concentrate on educating the Host SC and its support staff. Others believed that the one-week review, inspecting the competition material, was really needed, though it would have been better to conduct it three months before the IOI rather than two months. 4. IOI'01 and beyond We discussed some of the expected changes for IOI'01 in Finland and how to prepare for them. In particular, we considered * Linux operating system: - its developer, Linus Torvalds, is from Finland - advantages over MS-DOS, it is a modern O.S., especially stable (no rebooting after failed test runs), multitasking, long file names, secure file system, networking support, better timing (process based), It is not clear whether Linux-only is feasible. * New compilers for C/C++ and Pascal. Rob Kolstad has done most of the work on this and will report separately. Possibly we can set up one computer in the GA computer room with the new compilers. * Non-secret input data, especially when solutions involve heuristics. Not likely to happen at IOI2000, but feasible for IOI'01. * More test cases per task. * I/O always (both for batch and reactive programming tasks) via stdio or always via a library. * Source code submission. Evaluation would still based on black-box testing. * External submission of tasks. Requirements, confidentiality, time-line, who finishes the work? * Role of ISC. There are still some things to work on, such as refining the Guidelines for IOI Competitions, and applying them to more aspects of the organization. We made no final decisions for any of these items. They will be brought to the attention of the GA at IOI2000. 5. General No further items came up. 5. Closing Tom closed the meeting.