ISAC: Minutes of ISAC meeting held on 25 May 2006

Minutes of ISAC meeting held via telecom on 25 May 2006

Present: Peter Ashenden, Larry Soule, Chuck Swart, Lance Thompson, Ajay Verikat

Absent: Jim Lewis

Next Meeting: July, perhaps at DAC

TOPIC: Timing with respect to Accellera LRM

If we can complete ISAC analysis/approval and get rapid VASG approval
(by June 2) then we can incorporate IR updates into the proposed
LRM. The LRM will be completely frozen by June 5.

TOPIC: IR 2082 Elaboration of unconstrained interface objects

This was ISAC-approved.

ACTION: Chuck to forward to VASG chair.

TOPIC: IR 2085 What happens when a parameter of mode out is not
assigned in a procedure?

This was ISAC-approved. The analysis is accepted, the LRM rewording is
covered in IR 2082.

ACTION: Chuck to forward to VASG chair.

TOPIC: IR 2092 Type conversions don't allow for null arrays

The basic analysis was accepted. However it was agreed that there
should not be a blanket restriction against enumeration types with
only one element. Such types could be created as placeholders to be
added to as a project advances. Instead, it should be an error to
create a null string if the base type of the index subtype does not
have at least two elements.

ACTION. Chuck to rewrite, including an example. All to vote.

TOPIC: IR 2038 Minor semantic errors

This long, detailed IR was ISAC-Approved. Good job Ajay!  As part of
the analysis the issue was raised that the LRM describes the subtype
of an array aggregate, which, technically should have index ranges but
no subtype. It was agreed to temporarily let this issue go, and to
later write up an IR to examine all references to type, subtype, base
type to make sure that they are all accurate.

ACTION: Chuck to  forward to VASG chair.

TOPIC: IR 2093 Static type conversions and qualified expressions

The IR was revised to not allow unconstrained array types to be
locally static. Related changes were made to constant declarations.
This was ISAC-Approved.

ACTION: Chuck to  forward to VASG chair.

TOPIC: IR 2094 Attribute specifications of overloaded subprograms is limited

Peter suggested an LRM change which allows a use case which is
currently forbidden, without affecting backward compatibility or
useful error checking.  This was ISAC-Approved.

ACTION: Chuck to forward to VASG chair.

TOPIC: IR 2095 What is the entity class of an enumeration literal?

It was agreed that the entity class of an enumeration literal should
be "literal" and not "function". Lance suggested a slight change to
section 3.1.1 to clearly indicate this. It was agreed that a note
would also be added to section 7.3.1.  This was ISAC-approved.

ACTION: Chuck to update, forward to VASG chair.