Re: [vhdl-200x] Voting Results: ISAC Response

From: Jim Lewis <jim_at_.....>
Date: Tue Apr 18 2006 - 11:11:02 PDT
Hi,
The following is the ISAC response to the vote that
closed on 3/20/06:

---------------------------------------------

In a recent VASG ballot the following IRs were voted on:

2075    Arrays with numeric and enumeration index types are not closely related
2080    Case expression should include parenthesized expression
2081    The term ancestor is used where parent is meant

IRs 2080 and 2081 were approved with no dissenting votes and no comments.

IR 2075 received 8 affirmative votes and 2 negative votes.  The
comments accompanying the negative votes were:

"Type conflict and ambiguity of elaborate counting schemes detracts
from the intended clarity of a) strong typing, b) sub-typing, c) clear
and unambiguous interfaces."

and

"I'm okay with making no change to 2002. In my opinion it's not worth
the effort to fix it in a future revision. The current definition,
although overly restrictive, is simple and clear."

The response of the ISAC is the following:

Strong typing is a major strength of VHDL. However, some mechanism is
necessary to provide a "change of view" of an object. This capability
is achieved in part by means of type conversions. If any part of the
language should relax strong typing, it should be in the area of type
conversions, which provide a controlled mechanism for changing types.
Sometimes a type conversion also provides a change in value, as in
conversions between integer and real types. So it does not seem
unreasonable to allow some form of conversions between enumeration
types and integer types when used as array indexes. Although it is
true that the method of determining the mapping between integers and
enumerations for unconstrained target types is arbitrary and
"elaborate", the proposed change has the strength that it maximizes
the cases in which type conversions succeed, similar to current rules
for concatenation.  Finally it is relatively easy for the implementing
tool to do this conversion and it is not trivial for the user to
perform the conversion. In summary, it seems that the downside is
small and that the increased ease given to the code developer is
substantial. Since there was such strong consensus in the vote the IR
is VASG-Approved.
Received on Tue Apr 18 11:11:06 2006

This archive was generated by hypermail 2.1.8 : Tue Apr 18 2006 - 11:11:34 PDT