Minutes:	IOI Scientific Committee Meeting
Status:		Approved (November 1999)
Time/place:	IOI'99, 15th October 1999, Antalya-Belek, Turkey

Present:
  Mohammad Ghodsi       Elected
  Gyula Horvath         Elected
  Jyrki Nummenmaa       Host n+1
  Ismail Hakki Toroslu  Host n-1
  Tom Verhoeff          Elected (chairman)
  Wenhu Wu              Host n
  Baolin Yin            Host n

  Alexandre Cerveira    (observer until chairman elected)
  Richard Forster       (taking minutes)


1. The meeting was opened by Tom Verhoeff, who pointed out that the ISC
is required to meet at every IOI.  A tentative agenda was present, along
with information about an ISC web page and mailing list at the secretariat.
Members were requested to verify their personal information.

2. In accordance with the ISC regulations a chair was elected.  Tom Verhoeff
was proposed by Mohammad Ghodsi and elected by general consent.

3. The question of the ISC's role at an IOI and during GA meetings was
raised.  It was the committee's understanding that ISC members would be
invited to an IOI and allowed to address the GA.  Richard Forster offered
to put this directly to the IC.

4. There was general agreement that ISC communications should be secure,
although this was less important for general (non-problem) discussion; PGP
was suggested for email (Tom Verhoeff will distribute information on PGP).
Other mechanisms, such as secure directories and physical media, were
suggested for larger material.  Each member will be responsible for keeping
their copies of all material secure.

5. Trust is an important issue for the committee.  Although bound by
the regulations the committee felt it could make a statement to the GA
that they would live up to the `vow of regulations'.

6. The meeting 2-3 months before an IOI was discussed.  All members were
hopeful that, if this was a physical meeting, they would be able to attend.
It was also suggested that the ISC, or some subset of its members, should
arrive that the IOI a few days early.  Baolin Yin will contact Zide Du
regarding host costs for China.

7. It was suggested that documentation should be available to the ISC
before the 2-3 month meeting and sent to any members who could not attend.
It was pointed out that it was not necessary for all members to receive the
same information at the same time; this would produce fresh perspectives
on changed documents.

8. Ismail Hakki Toroslu outlined the two part scientific team for IOI
'98.  There was a `Preparation' sub-team who produced the questions and
developed test data.  The tasks were then passed on to the `Auditing
and Writing' sub-team.  There was a continual implementation /
improvement cycle.

   It was asked if the ISC was just auditing and writing, or whether it
could suggest tasks.  Jyrki Nummenmaa said Finland would be in favour of
task suggestions and Baolin Yin said Chain was open at this stage and
would look at any suggestions.

9. The ISC would produce guidelines for what it expected to be delivered,
as well as a corresponding time-line.  It was suggested that, as well as
covering questions, this would include solutions and `proofs' for tasks,
evaluation information and test data, and information on alternate (i.e.
non-perfect) solution's performance and grading.  This would enable the
ISC to answer technical questions at GA meetings.

10. The guidelines might be peculiar to a given task type (e.g. heuristic).
Since the host SC needs to work to these guidelines this is a priority,
especially with regard to any new task types.  Committee members will
submit a criteria lists by mid-November, so that these can be consolidated
by the end of November.  It was agreed that Tom Verhoeff would discuss
this with the forthcoming hosts and then confer with the other committee
members.

11. Members were reminded that any new task types have to be submitted
before the IC meeting in February.

12. Heuristic tasks were not ruled out by either China or Finland.  Baolin
Yin said that all problems would have definite solutions in China.

13. Side events, such as tournaments, were floated.

14. The grading system and IST should be established as soon as possible.
It was noted that this is a long term issue and its time-line can be kept
distinct from that for IOI 2000, especially since the IOI platforms would
potentially change.  Rob Kolstad had been approached and expressed a
willingness to run the IST; Tom Verhoeff will approach him formally.  It
is believed that the IST will run using volunteers.

15. The roles of adjuncts and IST members needed to be kept separate, due
to the `no contact' requirement for adjuncts.  In particular, Rob Kolstad
would be approached for the IST and not as an adjunct.

16. A letter from Greece was submitted, although not discussed, regarding
the number of problems per day.

17. Meeting closed.