Minutes of ISAC meeting held via telecom on 
07 October 2004.

Present: Peter Ashenden, Jim Lewis, Deepak Pant, Chuck Swart, Ajay Varikat
Absent: (none)

Next Meeting: 20 October 2004. 8 pm Pacific Time
NOTE: Day change--Wednesday in USA Thursday in Asia/Oceana

TOPIC: IR2028 : Clarify Simulation Cycle

No new work done.

TOPIC: VHPI

Several people have reviewed Peter's document describing changes to
the simulation cycle.

ACTION: Chuck to forward this review to Peter.

TOPIC: IR2039: Minor typos

Peter supplied analysis.

ACTION: All to review, Chuck to issue call for email ballot.

TOPIC: IR2040: Problems with OTHERS in aggregates

There was general agreement that OTHERS should be legal in port maps
in which the formal contains a slice with a locally static range.
However,the way the current LRM is worded makes it difficult to incorporate
this change.
One way to resolve this problem is to give a subtype to a slice.
However,the effects of this are not yet completely understood.

ACTION: Ajay to attempt to write up both with a nd without the assumption
that slices have subtypes.

ACTION: Chuck to contact Doug Dunlop, who has considerable experience in
this area.

TOPIC: IR2041: Association of members is too restricted

This IR deals with the fact that individual association of OPEN is not
allowed for composites.  There is a technical problem with INPUT ports
and their default expressions.  For example, suppose that the port
declaration is an unconstrained bit-vector with initial value
"0000". Subelement association is used for subelements 1,2, 3,4 and
6. Subelement 5 is OPEN. How is this value determined?

It was suggested that since the current LRM is clear on this issue,
this should be considered an enhancement request and forwarded to the
M&P group

ACTION: Chuck to rewrite analysis with this perspective.

TOPIC: IR2042: Architecture as a block causes problems

Peter explained that purpose for treating an architecture as a block
within an entity was to give a was to declare the architecture name.
However, it was agreed that there are several undesirable effects.
Since peter has examined these issues in the context of VHPI he agreed
to supply the analysis.

ACTION: Peter to analyze.

TOPIC: FastTrack 26 Issue: Should subtype conversions be done on
qualified expressions?

FastTrack 26 deals with subtype conversions on return values from
functions. The above question came up during discussions of this
issue. It was decided that since resolution of this issue was not
essential for FastTrack 26, we should leave FastTrack 26 unchanged.

TOPIC: Web-based IR system.

Peter's proposed mapping was accepted, with one minor change.  Peter
has updated the mapping and has sent it to me.

ACTION: Chuck to look into implementation details, difficulty of
installing and supporting Bugzilla, etc.

ACTION: Chuck to forward revised mapping to ISAC.