We had a meeting of the Mercury group on Fri August 3, 2001.

Attendees:
	Zoltan Somogyi (zs)
	Mark Brown (dougl)
	Simon Taylor (stayl)
	David Jeffery (dgj)
	David Overton (dmo)
	Kevin Glynn (keving)

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 spent most of the week working on
		  teaching-related tasks.
		- Has finished implementing ignore counts in mdb.
		- Has improved the error messages for overloaded type
		  errors.
		- Has been working on stack slot optimisation.

	Mark Brown (dougl):
		- Has been working on the runtime system for retry in
		  .Net:
		  	- Has decided to use C# as much as possible
			  and only use MC++ for the low level
			  primitives.
			- Has almost finished the C# part.
		- Has been working on the garbage collection algorithm
		  for the log.

	Simon Taylor (stayl):
		- Fixed bug in smart recompilation tests.
		- Has submitted a diff for constraint propagation.
		- Has been working on moving mmake functionality into
		  the compiler.

	David Jeffery (dgj):
		- Has been working on HAL.
		- Fixed a bug in the option to turn on execution of
		  init code.
		- Is finishing up with HAL this week and has 2 weeks
		  working on his thesis before going overseas.

	David Overton (dmo):
		- Has been working on a paper on the constraint-based
		  mode analysis.
		- Has implemented a proposal by Ralph Becket to make
		  using higher order functions in standard modes
		  easier.  We discussed whether this was a good idea.
		  It was agreed that it was, but that it should wait a
		  few days to give other people the opportunity to
		  comment.

	Kevin Glynn (keving):
		- No work on Mercury again.
		- Mentioned that there will be a seminar next
		  Wednesday where Martin will be presented his work on
		  using constraint handling rules (CHR) for type
		  checking.

2. Other matters

	- Zoltan mentioned that Julien Fischer (juliensf@students),
	  the student who worked on the Java backend over summer, is
	  just starting honours and is looking for a Mercury-related
	  project.  He asked whether anyone had any ideas:

	  	- Mark suggested implementing Fergus's subtyping
		  proposal, but that it was thought it would be better
		  to wait until the type system is rewritten.

		- Another idea was that he might be interested in
		  using CHR for re-implementing the type system.
		  Kevin will make sure he gets sent the invitation to
		  Martin's seminar next week.

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