Minutes of ISAC meeting held via telecom on
02 September 2004.
Present: Deepak Pant, Chuck Swart, Ajay Varikat
Absent: Peter Ashenden, Jim Lewis

Next meeting: 16 September 2004.

TOPIC: IR2029 Non-relevant words and paragraph
Approved after final discussion.
ACTION: Chuck to forward to VASG

TOPIC: IR2031 "mod" function needed for TIME
Approved after final discussion.
ACTION: Chuck to forward to VASG

TOPIC: IR2032 Function "now" is not pure
Approved after final discussion.
ACTION: Chuck to forward to VASG

TOPIC: IR2036 protected_type_declarative_item includes subprogram_specification
Approved after final discussion.
ACTION: Chuck to forward to VASG

TOPIC: IR2040 Problems with OTHERS in aggregates
Discussion: The author raises three issues:
1.An OTHERS choice is allowed for formal parameters or formal generics but
not for formal ports. This was surely an oversight.
2. item a) states "a constrained array type (or subelement thereof)"
3. The formal is required to be declared with a constrained subtype, which
seems overly restrictive.
The analysis of issues 2 and 3 revealed that the expression
"or subelement thereof" is not clear whether or not this refers to the
formal or the actual. The LRM should be rewritten to make this clear.
If we interpret the phrase to apply to the actual, then issues 2 and 3 are
easier to resolve.
ACTION: ALL to think about this.

TOPIC: IR2041: Association of members is too restricted
Discussion: The author complains of three problems with association
of members.
1. OPEN is not allowed with individual association
2. Subelements should not be required to be with each other.
3. There should not be a requirement of an association for each member.
There are some issues to be considered here.

a. For input ports, if some elements are OPEN and others are not, then the 
default expression must be evaluated and some of its subelements used, others
ignored. 
b. Allowing subelements to not be with each other may cause subtle user errors,
for example, when the type is unconstrained, so that subelement values 
determine the appropriate range.
c. If the restriction on 1 is kept, then the requirement on 3 must be kept.
d. It is likely that some analyzers accept some of these technically illegal
conditions.
e. It is possible that some upstream code-generating tools violate the
requirement that all subelements be with each other.
ACTION: ALL think about these issue and decide if there are other issues.

TOPIC: IR1044 Definition of 'HIGH and 'LOW in a null range
Approved after final discussion.
ACTION: Chuck to forward to VASG