Re: Minutes of the Verilog-AMS meeting: 10th July 2008

It was probably my fault that this was listed as low-priority;
it seemed like a lot of work to track down the original intent,
and it didn't occur to me that copying the existing text to
Mantis would be a better way to mark the issue for later
attention than leaving it in the LRM.

I had a vague sense that maybe there were $functions mentioned
in scattered chapters, specific to the main topic of the chapter --
things like $param_given that is mentioned in 6.3.5 -- and
wondered if there were any $functions that were not also
described in Clause 9.  If an $identifier is defined outside
the list of places in 2.8.3, is someone going to be upset?

-Geoffrey



Bresticker, Shalom wrote:
> Hi, Sri.
> 
> The correct references are already there.
> The additional references are erroneous.
> 
> Regardless, there is no point in leaving references that are clearly
> wrong.
> If you are not going to change them, then delete them now, and you can
> always add others later. 
> 
> If you had written 2+2=5 and discovered the error, you would either
> correct it to 4 or delete it. You would not leave a sentence that you
> know to be wrong. It is different than a sentence which is partially
> correct and partially incorrect, where you might say that there is some
> benefit in leaving it.
> 
> Can you point to a single benefit from leaving the incorrect references?
> 
> Regards,
> Shalom
> 
>> -----Original Message-----
>> From: Sri Chandra [mailto:sri.chandra@freescale.com] 
>> Sent: Monday, July 14, 2008 6:09 AM
>> To: Bresticker, Shalom
>> Cc: Verilog-AMS LRM Committee; Wilmore, Jim
>> Subject: Re: Minutes of the Verilog-AMS meeting: 10th July 2008
>>
>> Shalom,
>>
>> Thanks for your response and clarifying the text on that.
>>
>> We had a bit of discussion on this item and we were not clear 
>> what the references are intended for, and what the correct 
>> references should be for both the system tasks and functions. 
>> It was acknowledged during the meeting that the references 
>> may be in error (due to chapter
>> addition/deletions) or sections having been moved. However, 
>> it was felt in the discussions that it was probably a minor 
>> issue which might take more time to actually figure out the 
>> correct ones and leave them there. 
>> May be not the best approach but since during the discussions 
>> it was felt as a minor item.
>>
>> I understand your concern, and greatly appreciate the 
>> feedback that we would like to incorporate in the LRM, and 
>> apologize that you find this particular decision ridiculous. 
>> I will take another look at this particular issue that you 
>> have mentioned.
>>
>> Regards,
>> Sri
>>
>>
>>
>> Bresticker, Shalom wrote:
>>> Hi,
>>>
>>> Regarding the cross-references to IEEE Std 1364-2005 claues 
>> for system 
>>> tasks and functions and compiler directives, the internal 
>> references 
>>> appear immediately following the texts in question:
>>>
>>> The $identifier system task or function can be defined in 
>> five places
>>> - A standard set of $identifier system tasks and functions, 
>> as defined 
>>> in Clause 8, Clause 10, Clause 17 and Clause 18 of IEEE std 
>> 1364-2005 
>>> Verilog HDL.
>>> - Additional $identifier system tasks and functions defined 
>> using the 
>>> PLI, as described in Clause 12 and Clause 20 of IEEE std 1364-2005 
>>> Verilog HDL.
>>> - Additional $identifier system tasks and functions defined 
>> in Clause 
>>> 4 and Clause 9 of this standard.
>>> - Additional $identifier system tasks and functions defined 
>> using the 
>>> VPI as described in Clause 11 and Clause 12 of this standard.
>>> - Additional $identifier system tasks and functions defined by 
>>> software implementations.
>>>
>>> and
>>>
>>> The `identifier compiler directive construct can be defined 
>> in three 
>>> places
>>> - A standard set of `identifier compiler directives defined 
>> in Clause 
>>> 11 and Clause 19 of IEEE std 1364-2005 Verilog HDL.
>>> - Additional `identifier compiler directives defined in 
>> Clause 10 of 
>>> this standard.
>>> - Additional `identifier compiler directives defined by software 
>>> implementations.
>>>
>>> Thus the references to Clauses 8, 10, 11, and 12 are 
>> clearly wrong and 
>>> it is ridiculous to leave them. The correct references do 
>> appear and 
>>> therefore these should be simply deleted. Even if you can 
>> claim that 
>>> maybe they were intended to refer to something else and you want to 
>>> find out what that was, it makes no sense to leave them in their 
>>> current form. Delete them now, and if you find in the future an 
>>> additional reference that should have appeared, add it 
>> then. As they 
>>> are now, they do not help anyone, and just confuse.
>>>  
>>>
>>>> * [Clause 2.8.3, pg 32]: The references to Clause 8, 10, and 12 of
>>>> 1364-2005 may be incorrect. This was discussed in the 
>> committee and 
>>>> its unclear at this point whether some of the clauses are internal 
>>>> references to Verilog-AMS document itself. Also chapter 
>> numbers have 
>>>> changed in p1364 and LRM2.3 and need to do detailed search for any 
>>>> reference to system task/function on these before removing these 
>>>> clauses.
>>>> ==> *Note:* This is not planned for this version and will be taken 
>>>> for next revision and also deemed as not very critical.
>>>>
>>>> * [Clause 2.8.4, pg 32]: The reference to Clause 11 of 
>> 1364 might be 
>>>> incorrect. This was discussed in the committee and its unclear at 
>>>> this point whether some of the clauses are internal references to 
>>>> Verilog-AMS document itself. Also chapter numbers have changed in 
>>>> p1364 and LRM2.3 and need to do detailed search for any 
>> reference to 
>>>> system task/function on these before removing these clauses.
>>>> ==> *Note:* This is not planned for this version and will be taken 
>>>> for next revision and also deemed as not very critical.
>>> Regards,
>>> Shalom
>>>

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