RE: [sv-ec] RE: [sv-bc] 5.15.3 Array reduction methods - result type

The problem is that "The Right Way" and "What the users want" don't
always agree (illegal drugs and Howard Stern-radio host, for example).

If you start breaking the rules for expressions, the opposite problems
will occur - people familiar with Verilog expression rules would expect
it to behave one way and it behave another way. There is also a problem
with this method breaking the rules for constraints, as it needs to be
treated as an operator and not a method to really be useful. (mantis
1517)

Dave


> -----Original Message-----
> From: owner-sv-bc@server.eda-stds.org [mailto:owner-sv-bc@server.eda-
> stds.org] On Behalf Of Bresticker, Shalom
> Sent: Friday, June 30, 2006 12:10 AM
> To: Steven Sharp
> Cc: sv-bc@server.verilog.org; sv-ec@server.verilog.org
> Subject: RE: [sv-ec] RE: [sv-bc] 5.15.3 Array reduction methods -
result
> type
> 
> But if it exists, it should be done The Right Way, according to what
> users need and want. Even more so if there are situations where there
is
> not another option. And Chris Spear tells me that users fall on this
> time and again.
> 
> > Note that you do have a way to work around the problem in many
> > situations:
> > stop using these methods and write a foreach loop to perform the sum
> > yourself, using an accumulator of whatever size you want.  I am not
> fond
> > of the idea of having built-in methods for things that are trivial
to
> > write
> > for yourself.  Apparently these methods were added primarily for use
> in
> > constraints, where that is not an option.
> 
> Shalom