Re: P1647 Initial Ballot Data

Dennis:

In your reply to Ron you mention:

>   With respect to your assertion of hidden agendas and villainy of those
> bent on Organizational Entity projects over Individual Expert projects to
> kill the DASC, the DASC was co-sponsor of several that have already been
> approved this year.  Just in case you are interested, we have become
> industry friendly in keeping with changes driven by the IEEE-SA.  If you
> think the DASC can only do  and has only done Individual Expert projects,
> you are wrong.

My question is (relevance of it discussed below):

1. What is ideal for DASC?
            A. To become more friendly to (a consolidated) EDA industry
            B. To become more friendly to the Semiconductor industry
            C. To optimize use of talented individuals to create
relevant standards
            D. To provide for a balanced process in which industry (both EDA
                 and Semiconductors) business forces, technical progress forces,
                 and competition regulatory forces act to produce
wealth enablers

You might guess that my answer is D.
What is wrong with A?  DASC friendliness may bring more money into the
process. True, for both the IEEE (those of you who won best paper
awards
might appreciate that), and for the groups which require support.
Entities meet, they decide what's best for next quarter
business, they even "agree to disagree" in a friendly way,
and "consensus" is reached in no time. Nothing wrong with that,
only positives so far...

However, there is a danger that any doc is quickly qualified as standard,
just as an enabler for selling a tool set.
Less high quality standards become acceptable, since the competition
regulatory forces are not present, and technical expert votes are
over-written
by one company (most of the time business driven) vote.
The lack of quality standards, costs more the Semi industry,
the frequency of their changes makes only the big survive, but best of
all the small guys are not able to compete because they do not have
any new
fast way to excel.  With no much progress made anyway,
the "run a standard wave revolution" is over.

After a few years people start realizing that the industry might be on
a plunging path, but some continue to claim that more corporate versus
sponsored individual participation in standard processes is needed in
order to fix the lack of progress.

The DASC was inexplicably forced to table the friendly proposal
to form a technical group to look at the existing standards, and how they relate
to new one to come (I remember a meeting in the IEC where Ron helped
craft that friendly proposal).
DASC is also placed in the weak position to just accept whatever
proposal for a new standard is submitted, instead of providing a list
of areas that require standardization, and requesting UNIQUE proposals
(as W3 Consortium does). Those are areas we have to address in order
to be able to prevent situations where,
      BAD_A Too many overlapping standards have to co-exist (like the
assertion situation today)
      BAD_B Standards lag behind the methodologies to produce modern ICs
         (i.e. no statistical approaches, DFM, ESL, reliable systems, 3D,
          EDA farm computing, etc.)
      BAD_C EDA Common definitions are not available let aside knowledge
        bases (some exceptions in PCB), and each new standard re-defines
        for example pin, port or direction
Why is all this relevant to our discussion? We seem to have people who
come from large corporations, who think that DASC should
go on an Entity path. We have others with a long experience of bringing
people together to craft progress who think that some other changes need to
occur to make DASC efficient and relevant at least at the initial level again.
For sure, we would not have any time to fill pages with what we think
about an alleged mishandling of information had we have to run ahead,
in fast gear, in order bring new tools to market, based on standards
that will be 100% relevant in, for example, 2008.

Finally, DASC ideal, if it helped only the EDA industry, could be to provide
standards that are complicated enough, that no one person alone would
be able to understand, and so, they would generate the need for more
and more tools and people to support them like, editors, rule
checkers, checkers that the assertions are correctly constructed etc.
I let you decide what is the role of individual contributors, and as
John would say their "ethics."

Any comments?

Sincerely,

Alex Zamfirescu