RE: [sv-ac] P1800: 1728 - formal proposal for Let Construct

Hi Ed,

I'll start by apologizing on taking so long to review this.  Here are my
comments:

16.6:  

a) there are two different paragraphs that compare the let to the
`define. Only one is needed.

b) do the restrictions on data types for assertions apply to let
(16.5.1)  I would think so.

c) The definition of "let_actual_arg" should probably include
"let_instantiation"

d) if "b" above is true, then should "strings" be allowed in
characteristic #11 since this is not a valid type for an expression? 

e) You have both modified the definition of expression to include
"let_instance" (16.6 and A.2.12) and created a new "expression_or_let"
(A.2.10).  These are redundant.  I think you need to only do the
expression_or_let to keep it from impacting non-assertions.

f) if I import a function from a package, and that function references
another function in that same package, are both functions automatically
imported, or just the one explicitely stated?

g) does the Annex A.12.2.10 at the top of the proposal need to show red
for key words and punctuation items?  Also, I think the A.12.2.10 should
be A.2.10 (see h below)

h) shouldn't A.2.12 be part of A.2.10 since it is for asseritons only?

i) shouldn't there be object diagrams for let in the chapter 36 since
there are VPI routines defined? I think existing object diagrams may
also be affected.

j) can "let" instances appear in action blocks?

k) are any formal semantics needed?

l) is there any talk of adding un-typed arguments to functions to get
around the issues of the type requirement?

Lisa




-----Original Message-----
From: owner-sv-ac@eda.org [mailto:owner-sv-ac@eda.org] On Behalf Of
Eduard Cerny
Sent: Tuesday, June 26, 2007 10:15 AM
To: sv-ac@eda-stds.org
Subject: [sv-ac] P1800: 1728 - formal proposal for Let Construct

Hello,

I have uploaded a new proposal for the let construct, this time in the
LRM format and written along the lines we discussed at the last meeting.
It is attached here for convenience.

Best regards,
ed


-- 
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.


-- 
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.