Re: [sv-bc] Mantis 2674 and 2611 (ballot 123 and 131)

From: Greg Jaxon <Greg.Jaxon_at_.....>
Date: Fri May 01 2009 - 16:21:20 PDT
So
integer T;
T x,y,x;
is not an error anymore? or do we just have to wait until elaboration to report it?

There is ample precedent for the scanner knowing whether a name will refer to an object vs some type.
These may be in the same "namespace" (unlike C/C++), but that doesn't mean that the scanner is unable to distinguish them.
Being in the same namespace just means that you cannot overload a name too much.
Compilation units (e.g., packages) were given their own namespace presumably so that you could overload
them with variable names!  I believe I've seen this in end-user RTL.

I object to having a long-standing feature of some implementations, about which no one has ever suggested
even a warning, suddenly become an error.

Greg

Francoise Martinolle wrote:
We do not want to treat type names differently than variable names.

Francoise
    '

--
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean. Received on Fri May 1 16:22:45 2009

This archive was generated by hypermail 2.1.8 : Fri May 01 2009 - 16:23:56 PDT