Re: [sv-bc] Proposal to make it easier to use packages with port declarations

Variables are obviously different; the original issue was about
including types.  Any non-matching type is observably different --
that is Brad's point.  Cases of class type inclusion are all non-matching.
My only point was that there are cases where inclusion of types
can produce matching types which is why I expressed concern
about using the term "different" when saying that type inclusion
produces "different" types -- the inclusion of a type into two
different compilation scopes MAY satisfy the "match" rules and thus
produce types that are not really "different".

Gord.

Arturo Salz wrote:

> An observable difference would occur when 'including files with variable
> declarations or class definitions that include static variables. In that
> case, "different" means that each scope would create its own version of
> those variables.
> 
> 	Arturo
> 
> -----Original Message-----
> From: owner-sv-bc@eda.org [mailto:owner-sv-bc@eda.org] On Behalf Of
> Gordon Vreugdenhil
> Sent: Wednesday, May 10, 2006 10:52 AM
> To: Brad Pierce
> Cc: sv-bc@eda.org
> Subject: Re: [sv-bc] Proposal to make it easier to use packages with
> port declarations
> 
> 
> 
> Brad Pierce wrote:
> 
> 
>>>Certainly for things such as unpacked structs
>>
>>
>>Regarding structs and unions, unpacked vs. packed does not matter
> 
> here,
> 
>>because two packed struct types do not match unless they were declared
>>with the same declaration statement.  (See 6.9.1.c.)  And the
> 
> difference
> 
>>between types that don't match is always observable (if one wishes)
> 
> via
> 
>>generate and the type operator. (See 6.10.)
> 
> 
> I didn't intend to be exhaustive, I just wanted to point out that
> using the term "different" is dangerous since in *some* cases
> "different" types (i.e. includes into compilation units) are not
> observably different.
> 
> Gord.
> 

-- 
--------------------------------------------------------------------
Gordon Vreugdenhil                                503-685-0808
Model Technology (Mentor Graphics)                gordonv@model.com