RE: [sv-ac] 1668 Clause 16 proposal updated

Ok,  I agree that it is very fine tuning and users will probably get the
right idea reading the text as is.

I wouldn't spend too much time on it

Doron

-----Original Message-----
From: John Havlicek [mailto:john.havlicek@freescale.com] 
Sent: Thursday, August 23, 2007 3:16 PM
To: Bustan, Doron
Cc: john.havlicek@freescale.com; sv-ac@eda-stds.org
Subject: Re: [sv-ac] 1668 Clause 16 proposal updated

Hi Doron:

I think that I understand your point.  I think that you are saying
something like the following:

   1. It can be deduced that if there are multiple semantic leading
      clocks, then the rules about the way local variables flow into
      composite sequences guarantee that there is a separate copy
      of the local variable for each maximal subproperty that has a
      single semantic leading clock.

   2. Therefore, we should not say that "Therefore, there are separate
      copies of the local variable".

I don't think the reader will generally understand this deduction, and
Dmitry has already suggested that we move the inductive definition of
the set of semantic leading clocks to Annex F.  I am now starting to
agree with him because the definition does not address all the derived
forms.  Doing so makes it even less likely that the reader will
understand the deduction.

I think that the writing in this part of the LRM needs to give the 
user the right idea, and it is contrasting the behavior of a local
variable declaration assignment in a property with several semantic
leading clocks with the case when there is only one semantic leading
clock.  

At the intuitive level, I'm not sure how important it is to decide 
whether the multiple copies are created by the flow into composite
operators or because there was a declaration assignment in the
presence of multiple semantic leading clocks.  The essential effect
should be the same.

I'm not entirely happy with your suggested rewrite, so I will try to 
think of a way to rewrite this that will satisfy us both.

J.H.

> X-ExtLoop1: 1
> X-IronPort-AV: E=Sophos;i="4.19,283,1183359600"; 
>    d="scan'208";a="265364111"
> X-MIMEOLE: Produced By Microsoft Exchange V6.5
> Content-class: urn:content-classes:message
> Date: Mon, 20 Aug 2007 09:36:42 +0300
> X-MS-Has-Attach: 
> X-MS-TNEF-Correlator: 
> Thread-Topic: [sv-ac] 1668 Clause 16 proposal updated
> Thread-Index: AcfirmeZ6IyJzDP/QJivgJC6tNVbHAARIRQw
> From: "Bustan, Doron" <doron.bustan@intel.com>
> X-OriginalArrivalTime: 20 Aug 2007 06:36:43.0578 (UTC)
FILETIME=[792601A0:01C7E2F4]
> 
> Hi John,
> 
> I review the proposals, the only minor comment I have is:
> 
> At the first paragraph after the example at page 8, I think that you
> should change:
> 
> Therefore, there are separate copies of the local variable v for the
two
> subproperties governed by these clocks.
> 
> To=20
> 
> Therefore, each copy of the two separate copies of the local variable
v
> for the two subproperties is governed by a different clock.
> 
> The two copies are there regardless of the different clocks.
> 
> 
> Doron
> 
> 
> -----Original Message-----
> From: owner-sv-ac@server.eda.org [mailto:owner-sv-ac@server.eda.org]
On
> Behalf Of John Havlicek
> Sent: Monday, August 20, 2007 1:13 AM
> To: sv-ac@server.eda-stds.org
> Subject: [sv-ac] 1668 Clause 16 proposal updated
> 
> Hi Folks:
> 
> I updated the Clause 16 proposal for 1668.
> 
> See the .doc file.
> 
> I can't attach it at the moment, so please go to
> mantis.
> 
> The only major change was to get rid of the=20
> restriction that sequences not match empty if
> a local variable declaration assignment is used.
> 
> J.H.
> 
> --=20
> 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.