Harry:
I am sorry, we are going in circles. I don't mind if one company leaves.
This was not the case for example with VHDL Analog. It too 9 years to
agree but no body left. The loosing proposal people were very upset,
rightfully so, but no one left the DASC.
Bill Hanna
-----Original Message-----
From: Harry Beatty III [mailto:jbeatty@us.ibm.com]
Sent: Monday, July 26, 2004 12:15 PM
To: stds-dasc@eda.org
Subject: RE: Review of DASC membership fees
John and Bill,
I do not support OZ's proposal to recognize CAG corporate
entities for a couple of reasons:
Corporations that join standards organizations are there for their
benefit. As soon as the organization is producing standards they
disagree with they leave and most likely join or start a competing
standards organization. For example there are several EDA standards
groups that compete with Accellera, one such organization is Si2. It
is
very easy to create a not for profit company with the purpose of
generating standards that are favorable to "Your company". In
contrast
the IEEE is held at a much higher level because it does not come
under
the same pressure as groups that have direct company memberships. Its
IEEE's open memberships that keep if from being dominated by the
worlds
largest companies, something I hope I never see.
Contrasting the IEEE standards development process with that of an
industry standards group such as Si2 or Accellera, the IEEE
development
process is much more level and thorough and as a result it often
takes
longer to complete. All of the standards bodies employ working
groups,
which are composed of interested members. Each produces a standards
document. The difference is in the process. Typically with an
industry
standards group the working group ballots its work amongst itself and
when approved by those that wrote it, it is submitted by the board.
When
the board approves it, it becomes "The Standard". In contrast the
IEEE
sends the ballot out to a much wider constituency. In the case of
1481
after the working group was satisfied they sent ballots to reviewers
from all over the world which was made up from a cross section of
industry, academia and individuals. After completing the ballot the
inputs from this cross section was analyzed and the document was
corrected and sent for re-ballot. After passing all of this there was
the RevCom review. While I'm sure someone can equate the IEEE steps
to
similar steps in an industry standards group's steps the diversity,
rigor and objectivity is, in my opinion, definitely no where near the
same level.
I view the standards community consisting of three tiers. The
first tier is the newly emerging world of "Open Source". This usually
occurs when a technical innovation becomes a commodity and is no longer
worth supporting or its so complicated that its not economically viable
for a single organization to pursue it. Linux is such an example. Here
source is the standard and no formal document is required. If you are in
doubt of the operation merely inspect the final product. The next tier
is the industry standards groups. These groups attempt to bring to
fruition a standard in a hurry that is beneficial to its community of
companies that make up its membership. The power of the membership
typically dictates the direction the standards it works on. Then there
is the international standards groups that deal with the broader
audience. The international groups represent the broadest spectrum of
users and to ensure everyone has a say they are very formal and very
time consuming. This is where the IEEE has excelled and I believe should
not compromise its standards of excellence. For the broadest possible
audience I believe IEEE should do more to encourage individuals and let
the companies be represented through those individuals.
Thanks,
John B.
Phone: 1-845-892-4212
Fax 1-845-892-2066
email jbeatty@us.ibm.com
|---------+---------------------------->
| | "John J. Shields"|
| | <jshields@ieee.or|
| | g> |
| | Sent by: |
| | owner-stds-dasc@e|
| | da.org |
| | |
| | |
| | 07/24/2004 12:08 |
| | AM |
| | Please respond to|
| | stds-dasc |
| | |
|---------+---------------------------->
>-----------------------------------------------------------------------
------------------------------------------------------------------------
|
|
|
| To: <stds-dasc@eda.org>
|
| cc:
|
| Subject: RE: Review of DASC membership fees
|
|
|
>-----------------------------------------------------------------------
------------------------------------------------------------------------
|
Thoughtful reply, Bill. Comments below.
From: owner-stds-dasc@eda.org [mailto:owner-stds-dasc@eda.org] On Behalf
Of Hanna, William A
Sent: Friday, July 23, 2004 1:57 PM
To: stds-dasc@eda.org
Subject: RE: Review of DASC membership fees
Importance: High
Hi John:
I don't see the points made by asking the DASC members to pay. DASC
average volunteer is a professional who is working on EDA tools or a use
of EDA tools. Standards benefit the industry that sprung aroung EDA
tools, namely the EDA industry. The EDA industry has to be involved for
two important
reasons:
1. Buy In, because if the EDA industry does not buy in, DASC future work
will not be seriously considered; the EDA vendor will develop own
de-facto standards;
2. By introducing new standards and/or standard extensions in a timely
fashion, the EDA industry that supported, or had an ibterest in the
Standard will have more potential for growth. Otherwise they stagnate
and stop to grow or even decline.
I agree that they do and they will buy in. In fact, they are placing a
priority on industry standards via Accellera, where they fund and drive
the results. You pay to play in Accellera. Some user organization see
it worth paying to be an Accellera stakeholder, too. This is not a
regressive step to fallback to defacto standards from one EDA vendor. It
is recognizing the value in getting consensus faster than the DASC
process does. Perhaps that will result in better tool sets. It has been
a wakeup call to the DASC to define its place EDA standards in the face
of Accellera and its own problems( with money, participation, pace of
progress, relevance of work, ...).
Once standards organization stop to produce worthwhile extensions, the
vendors will start pushing de-facto standards/extension, etc. This is
what I am referring to, going back to the dark ages of EDA. Because even
if the work of the vendors is good, the lack of universal applicability
makes it difficult to use tools from different vendors, what we here
refer to as Best In-Class.
I believe the vote will swing in favor of the customers who vote with
their pocket book by buying tools. That vote is not merely reactive to
defacto standards. Key customers make coordinated demands of the whole
tool supply chain in order to have a production worthy design flow. I
can agree with you that there is some likelihood of regressive behavior
by EDA vendors. I think the user community (the stakeholders, the
purchasers, not just some individuals with a technical interest) has to
push the vendors to fund open standards development. Is Accellera the
place or is the DASC or somewhere else. And, since it is your money
when you buy the tools, consider investing or pushing your company to
invest in protecting the standards process.
Otherwise, protect your money as a personal investment in DASC by
providing/fostering budget oversight. Don't let the DASC leadership
price you out of a role you personally want to play.
People like myself who are not employed directly in EDA activities will
stop interacting with the DASC. This already is very evident in the
present membership of the DASC that is for the most part EDA with very
little users (not EDA employed) types of members.
I am very concerned about this, and I wish I knew how to improve it. On
the other hand, I have seen it backfire where some users who do not have
the expertise that goes with being employed in the EDA industry,
contribute misguided requirements (or solutions) that lead to
development of features in EDA tools that have little value or rejection
of the "standard" by the industry leaders. In my final analysis, I want
a blend of individuals and entities from the industry and user community
involved. That is why I support Oz's suggestion to recognize CAG
corporate entities.
I also want a process that runs more reliably. I know it takes a
larger budget. As an individual, I want to protect my money and improve
the DASC's effectiveness, so I am willing to pay "a little" more. (Need
I say that I would prefer to pay less and support all initiatives to
keep the cost the same or lower it?) As a startup entity, I am cash
poor. I do not see an important enough stake to spend my money on an
entity contribution yet.
Bill Hanna
-----Original Message-----
From: John J. Shields [mailto:jshields@ieee.org]
Sent: Friday, July 23, 2004 4:30 PM
To: stds-dasc@eda.org
Subject: RE: Review of DASC membership fees
Hi Bill,
That is awfully dramatic. This is a more mature industry than
back
in the day when vendors
locked customers to proprietary tools indiscriminately. It is not
just customers like yourself
that have excellent institutional memory. There are many
partnerships between the customer
base and EDA vendors to get the whole job done. Standards have a
role where consensus
over time has more importance than a proprietary choice. It will
survive and individuals and entities
from both industry and user organizations will participate.
Oz has a well-reasoned opinion, whether you disagree with it or
just
his conclusion about user fees.
I agree with not creating a DASC entity class vs. co-opting the SA
corporate membership class.
He is also right in my opinion about all DASC members being
stakeholders and seeing value
in that membership. IF my membership goes up as proposed in order
to
demonstrate that
the stakeholders are doing their best to self-fund the agenda they
are setting, I believe the
debate should be agenda and the expected cost of it. I'm fine
with
the agenda, and I believe
we must tighten the way DASC operates and that proposed budget
severely
...because it is my money.
We should be respectful about the interaction with the treasurer
but
scrutinize the derivation
of the projected funding. The budget wants to balanced with a
reasonable uncertainty. No
deficit spending if at all possible. If we exceed expectations,
keep
a rainy day fund or jump
on it and spend it J
Regards,
John Shields
From: owner-stds-dasc@eda.org [mailto:owner-stds-dasc@eda.org] On
Behalf Of Hanna, William A
Sent: Friday, July 23, 2004 12:27 PM
To: stds-dasc@eda.org
Subject: RE: Review of DASC membership fees
Importance: High
DASC Team:
This discussion is leading this volunteer organization to
extinction.
The purest never invented, never developed, and never produced.
You
need the money and the power/influence of the EDA industry. If you
loose that, then membership will dwindle, product (proposed
standards/updates) will go down, and no one will be een interested
in
using what you produce. We will go back to before the VHSIC/VHDL
standards when every vendor of EDA tools had his own product
supported by internal documentation that was intentionally done to
keep the other guy away from the unlucky customer they caught in
their Net of misleading adds and false promises. Lord Have Mercy!
WAHanna
William A. (Bill) Hanna
Boeing SDB Programs
Boeing Technical Fellow-Advanced Avionics
MC S500-3100 Phone: 636-925-4802
St. Louis, MO 63166-0516 Fax: 636-925-3327
E-mail: william.a.hanna@boeing.com
-----Original Message-----
From: Oz Levia [mailto:Oz.Levia@synopsys.com]
Sent: Friday, July 23, 2004 12:24 PM
To: stds-dasc@eda.org
Cc: stds-dasc-sc@eda.org
Subject: RE: Review of DASC membership fees
Evan, All,
On a light note, I am impressed with the attempts to figure
out
the intent and motives. If some of us worked at the right
profession the US security would be guarantied :)
On a lesser lighter note: there is no conspiracy. Peter's
suggested fee increase and entity membership format and the
'budget' I provided are not orchestrated to support each
other.
Also, DASC membership is what it is: a FACT, not a 'fact'. I
did not make things up. If you do not trust my data go look
at
the membership roster on the website.
Seriously now:
I think entity based working group and ballots are very
useful.
That is the best way to take business interest into
consideration when a standard is created. At the same time
individuals are essential mix of the technology of
standards.
Technical experts are required to ensure a standard is
feasible
and implementable.
I also think that DASC does need a realistic budget. Is it
$45K
or not? I agree that it is debatable but $0 is not the right
number. There are costs and it is unfair to ask some one
else
to bare the cost just because some of us don't want to. I
think
EVERY participant in DASC derives a direct or indirect value
from his or her membership. If they tell you otherwise they
are
not being honest with themselves. I think the question to be
asked is "what value?"
Another point (some what disjoint) is that, in my opinion,
while DASC should support entity participation and ballot,
we
don't need to have an entity membership class. I think doing
so
now will only put us in competition with SA corporate
membership. I would suggest that we co-opt the entity
membership of SA to our benefit and allow any and all
corporate
entity members to participate and ballot DASC sponsored
PARs.
In my view DASC has a value: it is a group of individuals
that
are knowledgeable about EDA and standards. We should
capitalize
on this strength.
Proposal:
Keep membership in DASC limited to individuals. Increase
individual fees (annual) to the range of $300-$500. We can
expect the number of participants to drop by at lease a 1/3
(70-110) or more. Even at the low end of the scale, DASC
membership dues will amount to some $20K a year. While this
is
not in line with the 'budget', I think we will have more
credibility asking for help (meeting rooms, equipment, etc)
when we have taken some steps to account for our real
expenditures. I would also suggest other classes of
membership:
5 year membership at 30% discount and life membership capped
at
$5000. DASC should have provisions to allow for membership
at
reduced cost for individuals that need such help.
At the same time, we should allow DASC to support (sponsor)
WG
that are entity XOR individually balloted provided that
entities are members of SC corporate.
-Oz.
-----Original Message-----
From: owner-stds-dasc@eda.org
[mailto:owner-stds-dasc@eda.org]
On Behalf Of Evan Lavelle
Sent: Friday, July 23, 2004 2:02 AM
To: stds-dasc@eda.org
Subject: Re: Review of DASC membership fees
Oz, thanks for the figures.
(1) Your budget assumes:
Corporations: ~16
Over 1B: 9
Over 100M: 3
Over 10M 4
Individuals: ~45
Even with these assumptions, and individuals paying $200,
you
only get
$34K, not $45K. With individuals paying $40, you only get
$27K.
(2) Who are the 9 companies with revenues over $1B? The
entire
EDA
industry is only worth $4B, and only *2* companies have
revenues at
about $1B. Your potential revenues, using Peter's suggested
subscriptions, seem to me to be *way* out.
(3) You make the point (at length) that DASC is dominated by
tool
providers, and this 'fact' is presumably used to justify the
proposed
fee structure. However, this is basically wrong; it's not
that
simple.
You only name 4 providers: the top 3 and Verisity. However,
the
presentation doesn't take into account that the 54 Verisity
members are
a single-interest group: they have joined as individuals, to
get 1647
through. If they'd started a year later, they'd have gone
through on the
CAG, and we'd have 1 or 2 Verisity members, not 54. If you
exclude
Verisity, you get a much clearer picture, which indicates to
me
that
DASC is essentially individual-driven: Synopsys 14, Cadence
17,
Mentor
8, other vendors 16, the rest of us 40.
(4) Not your area, I know, but nobody has yet, after what
seems
to be an
eternity of discussions, made a business case for entity
membership. Is
anyone even going to try? As I said in my last mail, it's
completely
pointless having this discussion until that case has been
made
and
accepted. If the DASC can make that case, then you can
charge
the
providers whatever you want.
Evan
Received on Mon Jul 26 10:28:26 2004
This archive was generated by hypermail 2.1.8 : Mon Jul 26 2004 - 10:28:39 PDT