Subject: Re: [sv-ac] R29a/b Optional or mandatory name for assertions/properties/assumptions.
From: Adam Krolnik (krolnik@lsil.com)
Date: Wed Sep 18 2002 - 14:35:32 PDT
Hi ovi1757;
>Remember that ovi1751 has a user-assigned name: "Tom".
Oops, another mail came in, ovi1757, ovi1751 is now Simon.
Wait, are you Prakash or Richard - oh man...
>Even though I gave all of those reasons why a user name
>is nice, it still seems to me that it's the job of the
>language to allow for it, and it's the job of any design
Some languages actually build a methodology into the language.
They would argue that some things prevented makes for a better
language. Eiffel fits in this category.
Remember, coding guidelines are a critique of their language.
>mandating more things to specify with
>potentially garbage names "assertion foo" makes
>no sense.
You obviously have not worked with some designers out there in
the industry, ovi1757. Have you not seen entire blocks made
with garbage names for signals? Have you not seen instance
names that look like garbage?
Regarless of the quality of the name, people are less inclined
to change a name once set. Thus an assertion with a garbage or
good name has one that can be used by tools, etc and will not
receive this error:
Error, there is no assertion at line 24.
I have not seen the complaint about adding assertions because they
require a name. I've seen complaints about adding assertions
mostly because they didn't believe they would help get their
design working.
Who envisions a tool [theirs or another] that uses names and will
make them optional?
Adam Krolnik
Verification Mgr.
LSI Logic Corp.
Plano TX. 75074
This archive was generated by hypermail 2b28 : Wed Sep 18 2002 - 14:39:02 PDT