INDUSTRY GADFLY: EDA Goes OJ 
Author Message
 INDUSTRY GADFLY: EDA Goes OJ

 [ Editor's Note: This poll and write-up were done *before* Cadence decided
   to go after Cooper & Chyan.  (I'm getting engineers writing me trying to
   change their responses to my original survey questions *because* of
   this new Cadence/CCT lawsuit.)  Sorry, the polls are closed! :^) - John ]


    /o o\  /  it's a FEATURE!"                              (508) 429-4357
   (  >  )
    \ - /             INDUSTRY GADFLY: "EDA Goes OJ"
    _] [_       (published in EE Times as "Trial By Media")

                     by John Cooley, EE Times Columnist

        Holliston Poor Farm, P.O. Box 6222, Holliston, MA  01746-6222

  As anyone who followed the O.J. Simpson trial can tell you, we Americans
love "Trial by Media."  And the EDA business is no different.   Now that
EE Times has covered both sides of the Cadence-Avant! lawsuit in minute
detail, here's what engineers responding to a recent Internet survey said.

  Overall, 311 engineers responded: 30 percent used P&R tools directly,
29 percent supported other engineers using P&R tools, and 56 percent
only used related EDA tools like synthesis.  (Yup, that's a 115 percent
total -- some engineers wear multiple hats.)

  To the question "TRUE or FALSE: As an engineer, I believe the American
legal system is generally capable of rendering justice in technically complex
lawsuits.", only 27 percent said TRUE.  "I worked on the AMD/Intel suit with
Intel and the whole process was a joke.  The judge was so technically
clueless the lawyers (also technically clueless) would spend 110 percent of
their time trying to dumb down the data and predict how the judge would
react."  "TRUE -- but I would not want to have to stake my livelyhood on
this, though.  I was once disqualified as a juror for a Speeding Ticket trial
because I knew how RADAR worked."   17 engineers referenced O.J. Simpson
and 5 mentioned the McDonald's scalding coffee lawsuit.  "Can you spell OJ?
Can you spell DNA?"

  To the question: "As an engineer, if I were the judge in the Cadence/Avant!
lawsuit, I (WOULD / WOULD NOT) grant Cadence's Sept. 11 request for an
injunction to prevent further sales of Avant! products that are alleged to
contain Cadence technology.", 13 percent of engineers said they WOULD, 66
percent said they WOULD NOT, and 21 percent said they didn't know.  "WOULD
NOT -- once granted, the company's dead.  There's no appeal after death."
"Cadence is a large mega-company bullying its way into marketshare.  They
should be ashamed!"  "Something smells fishy to me at Avant!"  "I haven't
seen enough to prove guilt."

  To the question "Because of this lawsuit, I am (MORE LIKELY / LESS LIKELY
/ UNCHANGED) to do business with Cadence.", 81 percent chose UNCHANGED,
16 percent LESS LIKELY, and 3 percent MORE LIKELY.  "UNCHANGED -- But I feel
disgusted.  It is not easy to switch CAD systems especially when you have
spent years building it up.  I would if I could.  Their tools are not
improving in terms of quality."  "UNCH.  I hate Joe Costello and Bill Gates,
but my company depends on products from both."

  With the same question put in Avant! terms, 65 percent chose UNCHANGED,
29 percent LESS LIKELY, and 6 percent MORE LIKELY.  "LESS LIKELY.  I don't
want to be stuck with a future support issue."  Only 12 percent of users
voiced interest in non-Avant!, non-Cadence P&R tools.  "I think this lawsuit
will finish long before Compass Pathfinder finishes routing my 100k gate
design."

  The biggest surprise came when I later compiled the 36 additional responses
from EDA employees.  This group was three times as likely (42 percent versus
the EDA users' 13 percent) to grant Cadence's injunction.  And EDA makers
were twice as likely (53 percent vs. 27 percent) to see the courts as
competent -- which helps explain why the EDA industry is so litigious!

----

John Cooley runs the E-mail Synopsys Users Group (ESNUG), is past president
of the User Society for Electronic Design Automation (USE/DA) and works as a
contract ASIC/FPGA designer.  He loves e-mail from fellow engineers at



Sat, 06 Feb 1999 03:00:00 GMT  
 INDUSTRY GADFLY: EDA Goes OJ

Quote:

>  To the question "TRUE or FALSE: As an engineer, I believe the American
>legal system is generally capable of rendering justice in technically complex
>lawsuits.", only 27 percent said TRUE.  "I worked on the AMD/Intel suit with
>Intel and the whole process was a joke.  The judge was so technically
>clueless..."

While I wouldn't quarrel with the sentiment that there is often a problem
there, it's worth remembering that some judges do their homework.  Judge
Debevoise, whose denial-of-preliminary-injunction opinion induced USL to
settle out of court with BSDI and UCB, clearly understood the issues in
detail -- that opinion was a masterful summary, surprising and impressive.
--
 ...the truly fundamental discoveries seldom       |       Henry Spencer



Sun, 07 Feb 1999 03:00:00 GMT  
 INDUSTRY GADFLY: EDA Goes OJ

Quote:

>>  To the question "TRUE or FALSE: As an engineer, I believe the American
>>legal system is generally capable of rendering justice in technically complex
>>lawsuits.", only 27 percent said TRUE.  "I worked on the AMD/Intel suit with
>>Intel and the whole process was a joke.  The judge was so technically
>>clueless..."

>While I wouldn't quarrel with the sentiment that there is often a problem
>there, it's worth remembering that some judges do their homework.  Judge
>Debevoise, whose denial-of-preliminary-injunction opinion induced USL to
>settle out of court with BSDI and UCB, clearly understood the issues in
>detail -- that opinion was a masterful summary, surprising and impressive.

Similarly, the judges in the Communications Decency Act case in
Philadelphia did a masterful job; their decision's text includes one
of the best non-technical explanations of the net and its services
(WWW, FTP, telnet, e-mail, Usenet, gopher, IRC, etc) I've seen.
Once they got the facts straight the decision became obvious.

I must agree with John, though, that in many cases the courts botch it.
But when I think about the differences, in both the BSDI/UCB/AT&T case and
the ACLU/EPIC/lots-of-folks vs CDA case, at least one side had a strong
interest in educating the court.  In the AMD/Intel cases, it seems to me
that the lawyers for both sides were quite happy to spread as much
confusion as possible, working to throw out the judge just when he finally
had a clue meaning starting over with a judge who knows nothing, and it
may have been in some cases that the lawyers were acting more in their own
interests (prolong the case, collect lots of $$$) than in the interests of
the companies they were supposed to be representing.

Maybe courts need third parties without an axe to grind to educate the
judges on things like (say, in the recent EDA litigation) the modern chip
design process, what placement and routing are, an outline of which
algorithms are in the public research literature and what software was
released by universities, and all the other background technical questions
that both sides can agree to.

Me?  I have no idea whether Avant! ripped off Cadence or not.

--
-- Joe Buck     http://www.synopsys.com/pubs/research/people/jbuck.html

Work for something because it is good,
not just because it stands a chance to succeed.    -- Vaclav Havel



Mon, 08 Feb 1999 03:00:00 GMT  
 INDUSTRY GADFLY: EDA Goes OJ

Quote:

>Maybe courts need third parties without an axe to grind to educate the
>judges on things like (say, in the recent EDA litigation) the modern chip
>design process, what placement and routing are, an outline of which
>algorithms are in the public research literature and what software was
>released by universities, and all the other background technical questions
>that both sides can agree to.

>Me?  I have no idea whether Avant! ripped off Cadence or not.

Ironically, that's supposed to be the function of the "experts" they bring
into cases (but these "experts" seem to be brought in based on how they'll
influence a case; not on how objective they are.)

Anyway, concerning the Avant!/Cadence lawsuit, the general view that I
seemed to be getting from the engineers that I polled was sort of the
same view that most Americans have about various Republican/Democratic
scandals that periodically break out -- that is, both sides are "dirty"
and this is just a temporary wrinkle favoring one side.  Very few people
feel that the courts actually provide justice because of how they can
be so easily skewed by hot shot lawyers that wheelbarrels full of cash
can buy.  Although I don't have polling numbers, judging from the letters
I got that were anti-Cadence after the news of the Cadence/Cooper & Chyan
lawsuit broke, I'd say that the mood has gone further against Cadence
overall.
                           - John Cooley
                             Part Time EDA Consumer Advocate
                             Full Time ASIC, FPGA & EDA Design Consultant

===========================================================================
 Trapped trying to figure out a Synopsys bug?  Want to hear how 4599 other
 users dealt with it ?  Then join the E-Mail Synopsys Users Group (ESNUG)!


     /o o\  /  it's a FEATURE!"                 (508) 429-4357
    (  >  )
     \ - /     - John Cooley, EDA & ASIC Design Consultant in Synopsys,
     _] [_         Verilog, VHDL and numerous Design Methodologies.

     Holliston Poor Farm, P.O. Box 6222, Holliston, MA  01746-6222
   Legal Disclaimer: "As always, anything said here is only opinion."



Fri, 12 Feb 1999 03:00:00 GMT  
 INDUSTRY GADFLY: EDA Goes OJ

Quote:

>The biggest surprise came when I later compiled the 36 additional responses
>from EDA employees.  This group was three times as likely (42 percent versus
>the EDA users' 13 percent) to grant Cadence's injunction.  And EDA makers
>were twice as likely (53 percent vs. 27 percent) to see the courts as
>competent -- which helps explain why the EDA industry is so litigious!

This result from the survey still kind of stumps me.  Why are EDA vendors
significantly more legalistic than their customers?  I can't seem to figure
this one out.  Any insights anyone?

                           - John Cooley
                             Part Time EDA Consumer Advocate
                             Full Time ASIC, FPGA & EDA Design Consultant

===========================================================================
 Trapped trying to figure out a Synopsys bug?  Want to hear how 4599 other
 users dealt with it ?  Then join the E-Mail Synopsys Users Group (ESNUG)!


     /o o\  /  it's a FEATURE!"                 (508) 429-4357
    (  >  )
     \ - /     - John Cooley, EDA & ASIC Design Consultant in Synopsys,
     _] [_         Verilog, VHDL and numerous Design Methodologies.

     Holliston Poor Farm, P.O. Box 6222, Holliston, MA  01746-6222
   Legal Disclaimer: "As always, anything said here is only opinion."



Sat, 13 Feb 1999 03:00:00 GMT  
 INDUSTRY GADFLY: EDA Goes OJ

Quote:

>>The biggest surprise came when I later compiled the 36 additional responses
>>from EDA employees.  This group was three times as likely (42 percent versus
>>the EDA users' 13 percent) to grant Cadence's injunction.  And EDA makers
>>were twice as likely (53 percent vs. 27 percent) to see the courts as
>>competent -- which helps explain why the EDA industry is so litigious!
>This result from the survey still kind of stumps me.  Why are EDA vendors
>significantly more legalistic than their customers?  I can't seem to figure
>this one out.  Any insights anyone?

It's just a guess, but quite possibly the USERS think the value of the
software is how much it costs to distribute, while the VENDORS of the
software think the value is in what it costs to write in the first place.

If you perceive high value in the sales chain, then source code is not
worth fighting over. If you perceive value in the source code, then any
company that even LOOKS LIKE they have their hand in the cookie jar is
deemed a threat to your (or the industry's) revenue.

I wonder about the "courts as competent" numbers. Precisely how was the
question asked? You seem to equate granting the injunction with seeing
the courts as competent... or at least you indicate that others might.
Did I read that wrong?
--

SRE

* * *  -  -  -  -  -  -  -  -  -  -  -  -  -  -  -  * * *

* * *  -  -  -  -  -  -  -  -  -  -  -  -  -  -  -  * * *
* ftp:  192.100.81.1   415-508-0500   fax: 415-508-0501 *
* * *  -  -  -  -  -  -  -  -  -  -  -  -  -  -  -  * * *

TRY THIS:
  echo '[q]sa[ln0=aln256%Pln256/snlbx]sb3135071790101768542287578439snlbxq'|dc



Sun, 14 Feb 1999 03:00:00 GMT  
 INDUSTRY GADFLY: EDA Goes OJ


Quote:
>I wonder about the "courts as competent" numbers. Precisely how was the
>question asked? You seem to equate granting the injunction with seeing
>the courts as competent... or at least you indicate that others might.
>Did I read that wrong?

Steve, here are the two question exactly as worded.  They were separate;
that is, I didn't interlink them.

  1.) "As an engineer, I believe that the American legal system is generally
       capable of rendering justice in technically complex lawsuits."
       Please answer TRUE or FALSE:___________

  2.) "As an engineer, if I were the judge in the Cadence/Avant! lawsuit,
       I ( WOULD / WOULD NOT ) grant Cadence's Sept. 11 request for an
       injuction to prevent the further sales of Avant! products that are
       alleged to contain Cadence technology."  Please choose WOULD or
       WOULD NOT:____________________

There were more questions about what engineers used as sources of info on
the case and how the case effected their views about doing future business
with Cadence, Avant!, and other P&R vendors that followed these two.  That
data point (concerning how EDA vendors and EDA users saw the courts) was
based on responses from just these two question (above) though.

                           - John Cooley
                             Part Time EDA Consumer Advocate
                             Full Time ASIC, FPGA & EDA Design Consultant

===========================================================================
 Trapped trying to figure out a Synopsys bug?  Want to hear how 4599 other
 users dealt with it ?  Then join the E-Mail Synopsys Users Group (ESNUG)!


     /o o\  /  it's a FEATURE!"                 (508) 429-4357
    (  >  )
     \ - /     - John Cooley, EDA & ASIC Design Consultant in Synopsys,
     _] [_         Verilog, VHDL and numerous Design Methodologies.

     Holliston Poor Farm, P.O. Box 6222, Holliston, MA  01746-6222
   Legal Disclaimer: "As always, anything said here is only opinion."



Sun, 14 Feb 1999 03:00:00 GMT  
 INDUSTRY GADFLY: EDA Goes OJ

John Cooley replied to his own posting:

Quote:


> >The biggest surprise came when I later compiled the 36 additional responses
> >from EDA employees.  This group was three times as likely (42 percent versus
> >the EDA users' 13 percent) to grant Cadence's injunction.  And EDA makers
> >were twice as likely (53 percent vs. 27 percent) to see the courts as
> >competent -- which helps explain why the EDA industry is so litigious!

> This result from the survey still kind of stumps me.  Why are EDA vendors
> significantly more legalistic than their customers?  I can't seem to figure
> this one out.  Any insights anyone?

I think this is the NIMBY factor in action.  Many people that don't like
litigation will immediately start being litigious when something close
and dear to them is threatened.  I think that the EDA industry would
naturally react this way towards a suit in their field.  I highly doubt
that they would have reacted with such a large deviation from non-EDA
folks if the lawsuit in question involved a Defense contractor (or another
highly technical subject) instead of an EDA company.


Sun, 14 Feb 1999 03:00:00 GMT  
 INDUSTRY GADFLY: EDA Goes OJ

Quote:

>John Cooley replied to his own posting:

>> This result from the survey still kind of stumps me.  Why are EDA vendors
>> significantly more legalistic than their customers?  I can't seem to figure
>> this one out.  Any insights anyone?
>I think this is the NIMBY factor in action.  Many people that don't like
>litigation will immediately start being litigious when something close
>and dear to them is threatened.  I think that the EDA industry would
>naturally react this way towards a suit in their field.  I highly doubt
>that they would have reacted with such a large deviation from non-EDA
>folks if the lawsuit in question involved a Defense contractor (or another
>highly technical subject) instead of an EDA company.

I (as an EDA person) feel similarly.  They are sensitive to a perceived
injustice that strikes close to home.  But notice the way they lean.
Those in this field seem to recognize the feasibility of the Cadence
claims.  If they thought that it was a case of a "big fish" trying to
squash the competition - another sensitive issue - the reaction would have
been quite different.  Most EDA companies are much smaller than the
average electronics company (their customers), so the average employee
is closer to the business issues.
Disclaimer:  I work in EDA but have no connection to either company
and no inside knowledge of the situation.

--
Sean P. Morley                                (tel) 201-236-3635
VP Product Engineering                        (fax) 201-236-3655



Mon, 15 Feb 1999 03:00:00 GMT  
 INDUSTRY GADFLY: EDA Goes OJ


Quote:


>>The biggest surprise came when I later compiled the 36 additional responses
>>from EDA employees.  This group was three times as likely (42 percent versus
>>the EDA users' 13 percent) to grant Cadence's injunction.  And EDA makers
>>were twice as likely (53 percent vs. 27 percent) to see the courts as
>>competent -- which helps explain why the EDA industry is so litigious!
>This result from the survey still kind of stumps me.  Why are EDA vendors
>significantly more legalistic than their customers?  I can't seem to figure
>this one out.  Any insights anyone?

Uh, yeah.  Can you say "small sample size"?  For a sample of 36 EDA
employees, the difference between the expected and actual percentages
is about 9-10 responses.  Throw in the fact that it's a self-selected
sample, and things get even worse.  So if EDA employees in general
hold exactly the same opinions as EDA users, but you also got responses
from a dozen or so Cadence/ex-Cadence employees who followed the
company line, you would get the kind of results you report here.

I think the sample size is too small to draw any meaningful
conclusions about whether or not EDA vendors really do hold
different views from EDA users on this issue.  That doesn't keep
people from spouting off about it, of course.  And there might
be some real differences - we just don't have enough info to be sure.
--

It is quite humbling to realize that the storage occupied by the longest
line from a typical Usenet posting is sufficient to provide a state space
so vast that all the computation power in the world can not conquer it.



Sun, 21 Feb 1999 03:00:00 GMT  
 INDUSTRY GADFLY: EDA Goes OJ

Quote:

>The biggest surprise came when I later compiled the 36 additional responses
>from EDA employees.  This group was three times as likely (42 percent versus
>the EDA users' 13 percent) to grant Cadence's injunction.  And EDA makers
>were twice as likely (53 percent vs. 27 percent) to see the courts as
>competent -- which helps explain why the EDA industry is so litigious!

>This result from the survey still kind of stumps me.  Why are EDA vendors
>significantly more legalistic than their customers?  I can't seem to figure
>this one out.  Any insights anyone?


Quote:
>Uh, yeah.  Can you say "small sample size"?  For a sample of 36 EDA
>employees, the difference between the expected and actual percentages
>is about 9-10 responses.  Throw in the fact that it's a self-selected
>sample, and things get even worse.  So if EDA employees in general
>hold exactly the same opinions as EDA users, but you also got responses
>from a dozen or so Cadence/ex-Cadence employees who followed the
>company line, you would get the kind of results you report here.

Only two Cadence employees and one Avant! employee responded (which didn't
change the EDA vendor stats significantly.)

Concerning your sample size is too small & self selection reasoning, from
my sometimes questionable memory of a statistics course a zillion years
ago as a sophmore engineering student, I remember that these two arguements
aren't factors here.  That is, the 36 EDA Vendors and 311 EDA Users both
used the *same* selection process, so comparisions *can* be made between
the two populations.  In addition, there obviously is some sort of real
difference in the way the two groups see the world on *this* issue
because of the strong similarities between these two populations (both
are computer literate engineers who know something about electronic
design software.)  The only distinction is that one group buys EDA
software while another group makes EDA software.  Not that much of a
disjoint difference.   It's not like comparing EDA users to, say...
left-handed, Laplander {*filter*}s.

                           - John Cooley
                             Part Time EDA Consumer Advocate
                             Full Time ASIC, FPGA & EDA Design Consultant

===========================================================================
 Trapped trying to figure out a Synopsys bug?  Want to hear how 4599 other
 users dealt with it ?  Then join the E-Mail Synopsys Users Group (ESNUG)!


     /o o\  /  it's a FEATURE!"                 (508) 429-4357
    (  >  )
     \ - /     - John Cooley, EDA & ASIC Design Consultant in Synopsys,
     _] [_         Verilog, VHDL and numerous Design Methodologies.

     Holliston Poor Farm, P.O. Box 6222, Holliston, MA  01746-6222
   Legal Disclaimer: "As always, anything said here is only opinion."



Sun, 21 Feb 1999 03:00:00 GMT  
 INDUSTRY GADFLY: EDA Goes OJ


Quote:

>disjoint difference.   It's not like comparing EDA users to, say...
>left-handed, Laplander {*filter*}s.

Oh, yeah?

What makes you think we can't be compared....

Arnim "Lefty Lulu" Littek,

lefty Lapland {*filter*} from Lithgow, Lithuania.
--

Actrix Networks Ltd.                            fax +64-4-499-1130
uucp/PPP/SLIP/BBS accounts                      tel +64-4-499-1122



Thu, 25 Feb 1999 03:00:00 GMT  
 INDUSTRY GADFLY: EDA Goes OJ


:>The biggest surprise came when I later compiled the 36 additional responses
:>from EDA employees.  This group was three times as likely (42 percent versus
:>the EDA users' 13 percent) to grant Cadence's injunction.  And EDA makers
:>were twice as likely (53 percent vs. 27 percent) to see the courts as
:>competent -- which helps explain why the EDA industry is so litigious!

:>This result from the survey still kind of stumps me.  Why are EDA vendors
:>significantly more legalistic than their customers?  I can't seem to figure
:>this one out.  Any insights anyone?


:>Uh, yeah.  Can you say "small sample size"?  For a sample of 36 EDA
:>employees, the difference between the expected and actual percentages
:>is about 9-10 responses.  Throw in the fact that it's a self-selected
:>sample, and things get even worse.  So if EDA employees in general
:>hold exactly the same opinions as EDA users, but you also got responses
:>from a dozen or so Cadence/ex-Cadence employees who followed the
:>company line, you would get the kind of results you report here.

:Only two Cadence employees and one Avant! employee responded (which didn't
:change the EDA vendor stats significantly.)

:Concerning your sample size is too small & self selection reasoning, from
:my sometimes questionable memory of a statistics course a zillion years
:ago as a sophmore engineering student, I remember that these two arguements
:aren't factors here.  That is, the 36 EDA Vendors and 311 EDA Users both
:used the *same* selection process, so comparisions *can* be made between
:the two populations.  In addition, there obviously is some sort of real
:difference in the way the two groups see the world on *this* issue
:because of the strong similarities between these two populations (both
:are computer literate engineers who know something about electronic
:design software.)  The only distinction is that one group buys EDA
:software while another group makes EDA software.  Not that much of a
:disjoint difference.   It's not like comparing EDA users to, say...
:left-handed, Laplander {*filter*}s.

Ok, I didn't realize the number of EDA users responding was so
big, and the number of Cadence employees responding was so small.
Given that, yeah, if you can count on the responses being representative
of their respective populations, you do have statistical evidence
of a difference between the two groups.  At least that's what the
formula for the difference of the means of two binomial distributions
from my old stat text seems to tell me.  (The differences in the
proportions are 4.5 and 3.2 standard deviations away from the expected mean
if the populations were identically distributed.)

I'm still concerned about the self-selection bias, though.  I'm
concerned that the group of EDA vendors who would respond to such
a survey may not be representative of EDA vendors in general in the same
way as the group of EDA users responding is representative of EDA users
in general.  That's because the question is closer to home for the
EDA vendors, and may affect some types of vendors more closely than others.
Given the small number of EDA vendor responses, even a small self-selection
bias might have large effects on the overall results.

--

It is quite humbling to realize that the storage occupied by the longest
line from a typical Usenet posting is sufficient to provide a state space
so vast that all the computation power in the world can not conquer it.



Fri, 26 Feb 1999 03:00:00 GMT  
 INDUSTRY GADFLY: EDA Goes OJ


Quote:



>:>The biggest surprise came when I later compiled the 36 additional responses
>:>from EDA employees.  This group was three times as likely (42 percent versus
>:>the EDA users' 13 percent) to grant Cadence's injunction.  And EDA makers
>:>were twice as likely (53 percent vs. 27 percent) to see the courts as
>:>competent -- which helps explain why the EDA industry is so litigious!

>:>This result from the survey still kind of stumps me.  Why are EDA vendors
>:>significantly more legalistic than their customers?  I can't seem to figure
>:>this one out.  Any insights anyone?

Putting aside the statistics and sample size discussions, consider the
folowing:

The EDA users have expressed their opinion about this area, and it
affects their views of the vendors and probably their buying patterns
(including not buying a product due to disgust with behaviour).

Unless the sample of EDA vendor respondents included CEOs and Prez's etc,
the opinions of the EDA vendor respondents have little to do with what a
company does in this area, as these people have no influence on how or
why or when one EDA company sues another.

"Favorite EDA CEO": dial, dial, dial ...
"EDA CO Employee": ring, ring, ring, ... pickup
ECE: Hi, this is ECE, how may I help you?
FEC: Hi, this is FEC, and I'm calling all employees individualy to
     ask their opinion.
ECE: OK!
FEC: Theirs a little company that has a product that does what one
     of our products does, only they are better, cheaper, more
     customer responsive, fewer bugs, better futures plan, more open,
     less fattening, cures cancer, ...
     What do you think we should do?
ECE: Sue 'em!
FEC: Great, thanks for your input. By the way, how did you enjoy the
     latest re-org, where when all was said and done, we still had the
     same people in charge, all the same projects were still being
     worked on, my sacred cow project got slightly morre focus, the
     reporting structure changed radically but the individual
     contributors for each project didn't really change much, and we
     spent 3 weeks in turmoil convincing every one that things would
     be much better?
ECE: I can't say how much I thought it was a great change!
FEC: OK. bye.
ECE: bye.



Fri, 26 Feb 1999 03:00:00 GMT  
 INDUSTRY GADFLY: EDA Goes OJ


Quote:
>Ok, I didn't realize the number of EDA users responding was so
>big, and the number of Cadence employees responding was so small.
>Given that, yeah, if you can count on the responses being representative
>of their respective populations, you do have statistical evidence
>of a difference between the two groups.  At least that's what the
>formula for the difference of the means of two binomial distributions
>from my old stat text seems to tell me.  (The differences in the
>proportions are 4.5 and 3.2 standard deviations away from the expected mean
>if the populations were identically distributed.)

>I'm still concerned about the self-selection bias, though.  I'm
>concerned that the group of EDA vendors who would respond to such
>a survey may not be representative of EDA vendors in general in the same
>way as the group of EDA users responding is representative of EDA users
>in general.  That's because the question is closer to home for the
>EDA vendors, and may affect some types of vendors more closely than others.
>Given the small number of EDA vendor responses, even a small self-selection
>bias might have large effects on the overall results.

Dave, again, I ain't no statistics wizard, but from my old college sophmore
class in the topic I remember how the proff said that self selection isn't
as flawed as it's commonly believed.  The reason why is that you test on
non-self selecting parameters.  For example, if you had people self-select
for if they liked the color "blue" as their favorite, it's perfectly
legit to break the "blue" lovers into male and female groups and then
use it to compare how men & women like of dislike eating spinach.  The
reason why is that their color preference has no correlation to what foods
they may or may not like.

My contention is that EDA makers and EDA users are far, far, far more
alike not only in training, but also in work environments, education,
 -- even race, gender, and pay scale distributions that it's fair to
test both samples from the Cadence/Avant! lawsuit survey to find
real diffences between these two groups as far as how they view lawsuits
in their industry.
                           - John Cooley
                             Part Time EDA Consumer Advocate
                             Full Time ASIC, FPGA & EDA Design Consultant

===========================================================================
 Trapped trying to figure out a Synopsys bug?  Want to hear how 4599 other
 users dealt with it ?  Then join the E-Mail Synopsys Users Group (ESNUG)!


     /o o\  /  it's a FEATURE!"                 (508) 429-4357
    (  >  )
     \ - /     - John Cooley, EDA & ASIC Design Consultant in Synopsys,
     _] [_         Verilog, VHDL and numerous Design Methodologies.

     Holliston Poor Farm, P.O. Box 6222, Holliston, MA  01746-6222
   Legal Disclaimer: "As always, anything said here is only opinion."



Tue, 02 Mar 1999 03:00:00 GMT  
 
 [ 15 post ] 

 Relevant Pages 

1. INDUSTRY GADFLY: EDA Goes OJ

2. INDUSTRY GADFLY "Why I Hate Wally"

3. INDUSTRY GADFLY: Synopsys Redeemed; Summit Rises

4. INDUSTRY GADFLY: "Barbarians At tHe Gate"

5. INDUSTRY GADFLY: SpeedSim's Three Dark Clouds

6. INDUSTRY GADFLY: The Fall(ing) VIUF 95

7. INDUSTRY GADFLY "Why I Hate Wally"

8. INDUSTRY GADFLY: Synopsys Redeemed; Summit Rises

9. INDUSTRY GADFLY: "Barbarians At tHe Gate"

10. MIT Entrepreneurship series: Semiconductor Track: EDA and Beyond: Funding and Industry Trends

11. tcl/tk use in the industry + a job to go with it

12. CLIPPER5.3 & CTGR.OJ

 

 
Powered by phpBB® Forum Software