We had a meeting of the Mercury group on Fri 18 May, 2001.

Attendees:
	Zoltan Somogyi (zs)
	Fergus Henderson (fjh)
	Mark Brown (dougl)
	Simon Taylor (stayl)
	David Overton (dmo)
	David Jeffery (dgj)

1. Progress reports

	We went around the table, with each person in turn discussing what
	they had been working on recently and/or what they were planning
	to work on.

	Zoltan Somogyi (zs):
		- Has almost finished writing the deep profiling
		  paper.

	Fergus Henderson (fjh):
		- Has been working on allowing mode-specific clauses.
		  There is still a bug with using this in the presence
		  of intermodule optimisation.
		- Has been looking at the issues associated with
		  linking with the dotnet backend.  (The IL assembler
		  can only produce .dll files not .obj files, and
		  recompiling every .il file when something changes is
		  slow.)
		- Has been looking at a bug with detecting multiple
		  instance declarations when using split C files or
		  dynamic linking for the library. The suggested
		  solution was to use c2init to check for multiple
		  instance declarations.

	Mark Brown (dougl):
		- Has been working on a new paper on declarative
		  debugging.
		- Has spoken to Simon Mei about finishing quickcheck.
		  Simon's work has now been reviewed and is ready to
		  be committed.  Mark will write a news item for the
		  web site.

	Simon Taylor (stayl):
		- Has been debugging smart recompilation.

	David Overton (dmo):
		- Has been debugging the new mode checker and the
		  boolean constraint solver.

	David Jeffery (dgj):
		- Has been working on a Hal paper which was submitted
		  earlier this week.
		- Has been helping Anthony Senyard with his problems
		  with using Mercury on Windows.

2. Other matters

	- It was decided that nondet pragma C code should be
	  deprecated since it does not work in all grades, and can be
	  replaced by a combination of Mercury code (using
	  mode-specific clauses) and det/semidet C code.  Zoltan will
	  document alternatives to using nondet pragma C in the
	  reference manual.

	- Mark and Zoltan received a rejection for their debugging
	  paper which was submitted to TOPLAS about a year ago.  The
	  reviewers suggested some major revisions to the structure of
	  the paper before it should be resubmitted.

Minutes taken by David Overton (dmo@cs.mu.oz.au).