RE: [sv-bc] E-mail Ballot: Respond by Wed Sep 05 8am PDT

Dave, 

> > 1. It is not in the usual "CHANGE-TO", red strikeout, blue addition 
> > format.
> [DR] I have been using this format for years. In a large 
> proposal, it is not useful to repeat the original text. Are 
> the strikeouts not red and the additions not blue?

Actually no. Most of the new text appears in red also.


> > 2. There is a difference in the meaning or use of the 
> string argument
> to
> > 
> > $test$plusargs and $value$plusargs. The argument to 
> $test$plusargs is 
> > searched for literally, whereas the argument to $value$plusargs is
> also
> > interpreted as a format string.
> > 
> > The wording in 1364-2005 for these two functions was chosen very 
> > carefully to be sometimes different and sometimes the same. 
> I see now 
> > that Stu made some changes in the merge. For example, the 1364-2005 
> > version had, "The $test$plusarg system function searches 
> the list of 
> > plusargs for a user specified plusarg_string". In the merge, this 
> > 'plusarg_string' was changed to 'string'. Probably in 1364-2005, we 
> > should have changed the subclause title as well to "17.10.1 
> > $test$plusargs (plusarg_string)".
> > 
> > But the changes proposed in this Mantis would change the description
> of
> > $test$valueargs without making the parallel changes on
> $value$plusargs.
> > 
> > It also makes the name of the $test$plusargs argument the 
> same as that 
> > of the first argument to $value$plusargs, whereas in 
> 1364-2005, it was 
> > deliberately chosen to be different.
> [DR] Unless you have a better suggestion, I will take those 
> changes out of this proposal so all the rest of the changes 
> can applied. I can create a new mantis item for this. (but 
> I'm not volunteering)

I think this would be best.

 
> > 3. The proposal also deletes the sentence which Mantis 988 
> added about 
> > ignoring leading nulls in the string. That may be ok if the 
> string is
> of
> > type string, but I am not sure that it can be left out if 
> it is of an 
> > integral data type, such as the classic reg vector.
> [DR] Now that these arguments are defined as strings, the 
> implicit conversion of an integral value to a string will 
> remove leading \0's.
> See 7.8

Where does it say that the argument is string type? The text in the
proposal is, "The user_string is an expression that is a string literal,
string data type, or an integral data type."

Regards,
Shalom

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