OT? Rant: Professional pride

Hi all.

Many years ago, I attended some conference where some guy who did
not have English as his "natural 2nd language" gave a presentation.
This guy was the only one I saw during a full week to actually keep
the time schedule. He wrapped up his talk at exactly 15:00 minutes,
leaving five full minutes for questions and discussions. No questions
were asked from the audience, possibly because of the presenter's
not quite perfect English diction. Not that I blame him for that, he
did a better job than I would have managed if I was told to give a
presentation in Russian, Chinese of Japanese in six months time.

anyway, the chairman, a slick elegant English academic of Oxbridge
grooming, who was very aware he was a slick English academic of
Oxbridge grooming, found the silence soewhat awkward, and asked,
in a way only a slick English academic of Oxbridge grooming can,

"Excuse me sir, but what is new about this work?"

The reaction of the presenter was as if the chairman had pretended to
shake his hand but hit him to the stomach instead. And we got a speedy
almost verbatim re-run of the 15 minutes talk, this time in two
minutes.

After the session I went up to the chairman and asked if he hadn't been

a bit rough with the poor guy. "Oh no," he said, "that is the *one*
question
everybody who gives a talk at a scientific conference should always
be prepared for. If you can't answer it, you should not give the
presentation in the first place."

It did make sense to me.

So I adapted it to my own version as "Excuse me sir, why do you think
this
method will work?" applied to engineering. True, the question is a bit
tough
if asked out loud, but I find it useful as a philosophical basis. If I
can not
formulate an answer to explain the intention of using some method or
algorithm, I don't know why I use it and should probably not be doing
the
job in the first place.

That's the philosophical basis I have used for a decade. Whenever I
have
had a saying in a project, I have used the philosophy to suggest
solutions
when I can, and explain why I could not find any solution when that was

the case. Lots of people don't like to hear that something can *not* be

done. So I have more than once pointed out that Niels Henrik Abel,
the only Norwegian mathematician of international fame, is most widely
known in the general public, for having proved that it is *im*possible
to
solve fifth degree equations by means of the "standard" algebraic
methods.
I found some of the reactions to be all out disturbing.

It follows as a trivial corrolary that one should stay away from
projects
where no path towards a solution exists, unless the objective of the
project is to find such a path. Basically, it is a matter of getting a
pre-project
feasibility study, which is part of basic project managment
craftmanship.

Imagine my surprise, even digust, when I discussed some projects with
som guy a few days ago. This guy said he was not the least interested
in
feasability studies. "They always conclude with that whatever can not
be done." Which is just about true, what underwater acoustics is
concerned.
So no one write feasability studies. And no one validate the methods
they come up with. Neither is good for business.

And this seems to be intentional. I cited one particular project, that
was
based on the following argument:

Use an active sonar to insonify the water. If target A is present, the
recieved echo will show property X, or

A => X     [implication]

in logical formalism. I do not contest the implication.

The project was based on analyzing the recieved signal for property X,
using this as an indicator for the target A. In doing this, one
implicitly
puts a constraint one can not defend:

A <=> X   [equivalence]

The equivalence means that only target A, and no others, induces
property X in the recieved data. Which is a very strong claim.
What needs to be done is to make sure there exist no other targets
in the sea that show property X. Since "property X" in this particular
case
is formulated as "the target reflects frequencies f1,f2,...,fN" there
is no
way of establishing the equivalence. It is a ridiculous claim to say
that
one and only one type of object reflects a set of specified
frequencies.

So I pointed this out. There are most probably several targets that
show property X,

A => X
B => X
C => X

and so on ad infinitum. So making an observation X can not be
linked to target A in any definite way, basically kicking the feet
away beneath the project in question. "Yep, that's logical" was
the answer. Sure, it is. Students are required to take a basic course
of elementary logic before being allowed to take any degree in
a university. But the connection to the real-world project in question
was not grasped.

I just don't get it. Somebody proposes and sells a project where
there is absolutely no chance of success, whatsoever. When asked,
they see the argument why, and agree with it, but still go on. Why?
Do they have no professional pride at all? Is it worth the carreer
and reputation of the junior engineers that are assigned to do the
grunt work? There is a peculiar pattern in these types of projects,
that the basic ideas behind the projects are never criticized when
they blow up. It is always the assigned engineer who is not up to
the task.

Oh well.

There must be material for at least a couple of doctoral theses in
psychology, somewhere inside here.

Rune

"Rune Allnor" <allnor@tele.ntnu.no> writes:

> I just don't get it. Somebody proposes and sells a project where
> there is absolutely no chance of success, whatsoever. When asked,
> they see the argument why, and agree with it, but still go on. Why?
> Do they have no professional pride at all? Is it worth the carreer
> and reputation of the junior engineers that are assigned to do the
> grunt work? There is a peculiar pattern in these types of projects,
> that the basic ideas behind the projects are never criticized when
> they blow up. It is always the assigned engineer who is not up to
> the task.

Several times I've seen problems which, when interpreted one way, have
no solution.  However, the statement of the problem is usually
ambiguous enough to allow for other interpretations.  This allows
"cheating" (re-scoping of the problem so that it is tractable).

Cheating in this way can be good fruit for research or innovation. :-)

Ciao,

Peter K.

-- 
"And he sees the vision splendid 
of the sunlit plains extended
And at night the wondrous glory of the everlasting stars."

 

Rune Allnor wrote:

[Skipped]

> 
> There must be material for at least a couple of doctoral theses in
> psychology, somewhere inside here.
> 
> Rune
> 

The point of the projects, conferences, etc. is not in creating 
something great or at least doing something sensible. The real point is 
that someone is having a good time for a while, that's all. Admit this 
point of view and the life will be easier to you :)

Vladimir Vassilevsky

DSP and Mixed Signal Design Consultant

http://www.abvolt.com

Rune Allnor wrote:
> Hi all.
> 
> Many years ago, I attended some conference where some guy who did
> not have English as his "natural 2nd language" gave a presentation.
> This guy was the only one I saw during a full week to actually keep
> the time schedule. He wrapped up his talk at exactly 15:00 minutes,
> leaving five full minutes for questions and discussions. No questions
> were asked from the audience, possibly because of the presenter's
> not quite perfect English diction. Not that I blame him for that, he
> did a better job than I would have managed if I was told to give a
> presentation in Russian, Chinese of Japanese in six months time.
> 
> anyway, the chairman, a slick elegant English academic of Oxbridge
> grooming, who was very aware he was a slick English academic of
> Oxbridge grooming, found the silence soewhat awkward, and asked,
> in a way only a slick English academic of Oxbridge grooming can,
> 
> "Excuse me sir, but what is new about this work?"
> 
> The reaction of the presenter was as if the chairman had pretended to
> shake his hand but hit him to the stomach instead. And we got a speedy
> almost verbatim re-run of the 15 minutes talk, this time in two
> minutes.
> 
> After the session I went up to the chairman and asked if he hadn't been
> 
> a bit rough with the poor guy. "Oh no," he said, "that is the *one*
> question everybody who gives a talk at a scientific conference should
> always be prepared for. If you can't answer it, you should not give
> the presentation in the first place."
> 
> It did make sense to me.
> 
> So I adapted it to my own version as "Excuse me sir, why do you think
> this method will work?" applied to engineering. True, the question is
> a bit tough if asked out loud, but I find it useful as a philosophical
> basis. If I can not formulate an answer to explain the intention of 
> algorithm, I don't know why I use it and should probably not be doing
> using some method or the job in the first place.
> 
> That's the philosophical basis I have used for a decade. Whenever I
> have had a saying in a project, I have used the philosophy to suggest
> solutions when I can, and explain why I could not find any solution 
> when that was the case. Lots of people don't like to hear that something
> can *not* be done. So I have more than once pointed out that Niels
> Henrik Abel, the only Norwegian mathematician of international fame, is 
> known in the general public, for having proved that it is *im*possible
> most widely to solve fifth degree equations by means of the "standard"
> algebraic methods. I found some of the reactions to be all out
> disturbing.
 >
> It follows as a trivial corrolary that one should stay away from
> projects where no path towards a solution exists, unless the objective
> of the project is to find such a path. Basically, it is a matter of
> getting a pre-project feasibility study, which is part of basic project
> managment craftmanship.

Then one out to say  that one should stay away from projects where no 
path towards a solution is known to exist.

> Imagine my surprise, even digust, when I discussed some projects with
> som guy a few days ago. This guy said he was not the least interested
> in feasability studies. "They always conclude with that whatever can
> not be done." Which is just about true, what underwater acoustics is
> concerned. So no one write feasability studies. And no one validate the
> methods they come up with. Neither is good for business.
> 
> And this seems to be intentional. I cited one particular project, that
> was based on the following argument:
> 
> Use an active sonar to insonify the water. If target A is present, the
> recieved echo will show property X, or
> 
> A => X     [implication]
> 
> in logical formalism. I do not contest the implication.
> 
> The project was based on analyzing the recieved signal for property X,
> using this as an indicator for the target A. In doing this, one
> implicitly
> puts a constraint one can not defend:
> 
> A <=> X   [equivalence]
> 
> The equivalence means that only target A, and no others, induces
> property X in the recieved data. Which is a very strong claim.
> What needs to be done is to make sure there exist no other targets
> in the sea that show property X. Since "property X" in this particular
> case is formulated as "the target reflects frequencies f1,f2,...,fN"
> there is no way of establishing the equivalence. It is a ridiculous
> claim to say that one and only one type of object reflects a set of
> specified frequencies.
> 
> So I pointed this out. There are most probably several targets that
> show property X,
> 
> A => X
> B => X
> C => X
> 
> and so on ad infinitum. So making an observation X can not be
> linked to target A in any definite way, basically kicking the feet
> away beneath the project in question. "Yep, that's logical" was
> the answer. Sure, it is. Students are required to take a basic course
> of elementary logic before being allowed to take any degree in
> a university. But the connection to the real-world project in question
> was not grasped.

Three local ministers were concerned that so many congregants came drunk 
to services, and set out to determine the cause. They spent one night 
drinking Scotch and soda, with predictable effect. Another night, they 
drank whiskey and soda to the same effect. On a third night, they drank 
Bourbon and soda, getting drunk again. Their conclusion was easy: soda 
is the cause of intoxication.

> I just don't get it. Somebody proposes and sells a project where
> there is absolutely no chance of success, whatsoever. When asked,
> they see the argument why, and agree with it, but still go on. Why?
> Do they have no professional pride at all? Is it worth the carreer
> and reputation of the junior engineers that are assigned to do the
> grunt work? There is a peculiar pattern in these types of projects,
> that the basic ideas behind the projects are never criticized when
> they blow up. It is always the assigned engineer who is not up to
> the task.

That's the unfair part. I worked in a research organization. I had to 
tell a new boss that it was perfectly all right to assign me a job with 
clear goals, tell me how it would be accomplished, and how long it will 
take. He had to realize, though, that the last two specifications were 
mere predictions. Moreover, Whether they proved accurate or not, the 
project could hardly be called "research".

> Oh well.
> 
> There must be material for at least a couple of doctoral theses in
> psychology, somewhere inside here.

The writer of the paper will have to differentiate speculative projects 
(those where no solution is known to exist but which seem worthwhile 
anyway) from the ones with asinine managers who, like a landlord I had 
in my early 20s, think that engineers can do anything.

Long ago, I needed a character generator for a dot-matrix printer. We 
know that ROM would be available in a year or two, but we needed the 
generator in the short term. I undertook build one with magnetic cores, 
threading each with sense wires representing the dots in the matrix for 
one letter, and with seven select wires, one for each ASCII bit. The 
half-select problem seemed insurmountable, though requiring unobtainable 
tolerances. My lab director at the time held the patents on memory cores 
and was the recognized expert in the field. (Jan Rajchman) We discussed 
the difficulties and he could see no way around them. He sensed my 
resignation as I was leaving his office and said to me as I approached 
the door that he wanted me to continue looking for a way: "If there is 
one, you will find it." In less than a week, I did. I think he felt even 
better about that than I did. (Tolerance didn't affect my solution. It 
easily accommodated +/- 30% on the select currents and any mix of core 
thresholds in the batch I had.

Later, another director assigned me to get a machine (of his design) to 
work without changing it. I was almost fired over that one, and would 
have been if he hadn't been fired well, moved upstairs) in my place.

One former director (former because he was so abrasive and sarcastic 
that few could work under him) had been promoted to assistant to the 
president. (In reality, general snoop and Grand Inquisitor.) What 
stopped him from being as good as he might have been at answering 
management's question, "How's it going?" was that everybody was afraid 
of him. If you were bullshitting, he would know before anyone else, 
maybe even yourself. He waltzed into my room one day -- I was trying to 
build a contactless keyboard with rollover (RCA still believed in 
computers then) -- and wanted to know how the project was coming. I 
answered his questions in a clear voice that hardly shook at all. Then 
he asked me a real tough one: "Why make this particular choice?" I 
figured that honesty was, at that moment, the worst of all possible 
choices, but conscience won out. I asserted that /some/ choice had been 
necessary, and this one seemed good at the time. He then asked what 
alternatives I had entertained, and I said, "None". He bought it.

We became, if not buddies, respected acquaintances. I devised ways to 
water his flowers while away without limiting flow with orifices so 
small they would clog. I figured out how he could legally stop the 
Pitot-static airspeed indicator on his plane from freezing up without a 
mechanic's license. We brainstormed about problems I wanted to crack; 
usually about work, but not always.

It seems that for all his bluster, what he wanted was someone who didn't 
flinch. (As a researcher, he had built the first power transistor -- one 
watt -- while a meeting was being held to discuss the feasibility of 
ever reaching a quarter watt. He had left the meeting impatiently and 
returned before it was over with a pellet soldered (by hand) to a copper 
plate. He chucked the thing still warm into the middle of the conference 
table and said, "One watt, ten minutes, still going strong when I 
disconnected it."

Jerry
-- 
Engineering is the art of making what you want from things you can get.
&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;

Vladimir Vassilevsky wrote:
> Rune Allnor wrote:
>
> [Skipped]
>
> >
> > There must be material for at least a couple of doctoral theses in
> > psychology, somewhere inside here.
> >
> > Rune
> >
>
> The point of the projects, conferences, etc. is not in creating
> something great or at least doing something sensible. The real point is
> that someone is having a good time for a while, that's all.

This is very true. There was a few years when I went to conferences
all over the mediterranean. Being a student, I had to fly with APEX
tickets, which "forced" me to stay  a few days longer in Greece and
Italy...

> Admit this
> point of view and the life will be easier to you :)

Yep. This is the point. I just wish somebody made that point twelve
years 
ago, so I could have chosen a different line of work.

Rune

Jerry Avins wrote:
> Rune Allnor wrote:
> > Hi all.
> >
> > Many years ago, I attended some conference where some guy who did
> > not have English as his "natural 2nd language" gave a presentation.
> > This guy was the only one I saw during a full week to actually keep
> > the time schedule. He wrapped up his talk at exactly 15:00 minutes,
> > leaving five full minutes for questions and discussions. No questions
> > were asked from the audience, possibly because of the presenter's
> > not quite perfect English diction. Not that I blame him for that, he
> > did a better job than I would have managed if I was told to give a
> > presentation in Russian, Chinese of Japanese in six months time.
> >
> > anyway, the chairman, a slick elegant English academic of Oxbridge
> > grooming, who was very aware he was a slick English academic of
> > Oxbridge grooming, found the silence soewhat awkward, and asked,
> > in a way only a slick English academic of Oxbridge grooming can,
> >
> > "Excuse me sir, but what is new about this work?"
> >
> > The reaction of the presenter was as if the chairman had pretended to
> > shake his hand but hit him to the stomach instead. And we got a speedy
> > almost verbatim re-run of the 15 minutes talk, this time in two
> > minutes.
> >
> > After the session I went up to the chairman and asked if he hadn't been
> >
> > a bit rough with the poor guy. "Oh no," he said, "that is the *one*
> > question everybody who gives a talk at a scientific conference should
> > always be prepared for. If you can't answer it, you should not give
> > the presentation in the first place."
> >
> > It did make sense to me.
> >
> > So I adapted it to my own version as "Excuse me sir, why do you think
> > this method will work?" applied to engineering. True, the question is
> > a bit tough if asked out loud, but I find it useful as a philosophical
> > basis. If I can not formulate an answer to explain the intention of
> > algorithm, I don't know why I use it and should probably not be doing
> > using some method or the job in the first place.
> >
> > That's the philosophical basis I have used for a decade. Whenever I
> > have had a saying in a project, I have used the philosophy to suggest
> > solutions when I can, and explain why I could not find any solution
> > when that was the case. Lots of people don't like to hear that something
> > can *not* be done. So I have more than once pointed out that Niels
> > Henrik Abel, the only Norwegian mathematician of international fame, is
> > known in the general public, for having proved that it is *im*possible
> > most widely to solve fifth degree equations by means of the "standard"
> > algebraic methods. I found some of the reactions to be all out
> > disturbing.
>  >
> > It follows as a trivial corrolary that one should stay away from
> > projects where no path towards a solution exists, unless the objective
> > of the project is to find such a path. Basically, it is a matter of
> > getting a pre-project feasibility study, which is part of basic project
> > managment craftmanship.
>
> Then one out to say  that one should stay away from projects where no
> path towards a solution is known to exist.

If the project has committed to deliver a working solution, then yes.
Most projects I have been involved with blew up because such
commitments were made, and no solutions could be found.

> > Imagine my surprise, even digust, when I discussed some projects with
> > som guy a few days ago. This guy said he was not the least interested
> > in feasability studies. "They always conclude with that whatever can
> > not be done." Which is just about true, what underwater acoustics is
> > concerned. So no one write feasability studies. And no one validate the
> > methods they come up with. Neither is good for business.
> >
> > And this seems to be intentional. I cited one particular project, that
> > was based on the following argument:
> >
> > Use an active sonar to insonify the water. If target A is present, the
> > recieved echo will show property X, or
> >
> > A => X     [implication]
> >
> > in logical formalism. I do not contest the implication.
> >
> > The project was based on analyzing the recieved signal for property X,
> > using this as an indicator for the target A. In doing this, one
> > implicitly
> > puts a constraint one can not defend:
> >
> > A <=> X   [equivalence]
> >
> > The equivalence means that only target A, and no others, induces
> > property X in the recieved data. Which is a very strong claim.
> > What needs to be done is to make sure there exist no other targets
> > in the sea that show property X. Since "property X" in this particular
> > case is formulated as "the target reflects frequencies f1,f2,...,fN"
> > there is no way of establishing the equivalence. It is a ridiculous
> > claim to say that one and only one type of object reflects a set of
> > specified frequencies.
> >
> > So I pointed this out. There are most probably several targets that
> > show property X,
> >
> > A => X
> > B => X
> > C => X
> >
> > and so on ad infinitum. So making an observation X can not be
> > linked to target A in any definite way, basically kicking the feet
> > away beneath the project in question. "Yep, that's logical" was
> > the answer. Sure, it is. Students are required to take a basic course
> > of elementary logic before being allowed to take any degree in
> > a university. But the connection to the real-world project in question
> > was not grasped.
>
> Three local ministers were concerned that so many congregants came drunk
> to services, and set out to determine the cause. They spent one night
> drinking Scotch and soda, with predictable effect. Another night, they
> drank whiskey and soda to the same effect. On a third night, they drank
> Bourbon and soda, getting drunk again. Their conclusion was easy: soda
> is the cause of intoxication.
>
> > I just don't get it. Somebody proposes and sells a project where
> > there is absolutely no chance of success, whatsoever. When asked,
> > they see the argument why, and agree with it, but still go on. Why?
> > Do they have no professional pride at all? Is it worth the carreer
> > and reputation of the junior engineers that are assigned to do the
> > grunt work? There is a peculiar pattern in these types of projects,
> > that the basic ideas behind the projects are never criticized when
> > they blow up. It is always the assigned engineer who is not up to
> > the task.
>
> That's the unfair part. I worked in a research organization. I had to
> tell a new boss that it was perfectly all right to assign me a job with
> clear goals, tell me how it would be accomplished, and how long it will
> take. He had to realize, though, that the last two specifications were
> mere predictions. Moreover, Whether they proved accurate or not, the
> project could hardly be called "research".

I don't agree with that. I agree with being assigned a goal and a
time/cost horizon. I do not, from consistent experience over many
years, trust anybody who tells me how to do things. If somebody
wants my cooperation, tell me what resources I have available,
and leave the rest to me.

Including time and resources for a feasability study to discover
if, and if so how, the goals can be achieved within the constraints.

> > Oh well.
> >
> > There must be material for at least a couple of doctoral theses in
> > psychology, somewhere inside here.
>
> The writer of the paper will have to differentiate speculative projects
> (those where no solution is known to exist but which seem worthwhile
> anyway) from the ones with asinine managers who, like a landlord I had
> in my early 20s, think that engineers can do anything.

That's the problem. I often wonder if anybody know the difference
between
a "witch doctor" and a "doctor of engineering", which is the direct
translation
of my academic degree.

> Long ago, I needed a character generator for a dot-matrix printer. We
> know that ROM would be available in a year or two, but we needed the
> generator in the short term. I undertook build one with magnetic cores,
> threading each with sense wires representing the dots in the matrix for
> one letter, and with seven select wires, one for each ASCII bit. The
> half-select problem seemed insurmountable, though requiring unobtainable
> tolerances. My lab director at the time held the patents on memory cores
> and was the recognized expert in the field. (Jan Rajchman) We discussed
> the difficulties and he could see no way around them. He sensed my
> resignation as I was leaving his office and said to me as I approached
> the door that he wanted me to continue looking for a way: "If there is
> one, you will find it." In less than a week, I did. I think he felt even
> better about that than I did. (Tolerance didn't affect my solution. It
> easily accommodated +/- 30% on the select currents and any mix of core
> thresholds in the batch I had.
>
> Later, another director assigned me to get a machine (of his design) to
> work without changing it. I was almost fired over that one, and would
> have been if he hadn't been fired well, moved upstairs) in my place.
>
> One former director (former because he was so abrasive and sarcastic
> that few could work under him) had been promoted to assistant to the
> president. (In reality, general snoop and Grand Inquisitor.) What
> stopped him from being as good as he might have been at answering
> management's question, "How's it going?" was that everybody was afraid
> of him. If you were bullshitting, he would know before anyone else,
> maybe even yourself.

What do you mean by "bullshiting"? Pretending you had answers you
did not? Doing useless work?

> He waltzed into my room one day -- I was trying to
> build a contactless keyboard with rollover (RCA still believed in
> computers then) -- and wanted to know how the project was coming. I
> answered his questions in a clear voice that hardly shook at all. Then
> he asked me a real tough one: "Why make this particular choice?" I
> figured that honesty was, at that moment, the worst of all possible
> choices, but conscience won out. I asserted that /some/ choice had been
> necessary, and this one seemed good at the time. He then asked what
> alternatives I had entertained, and I said, "None". He bought it.

That's the key point of mys tarting anecdote: SOME choise have to be
made. If gadget A is what you have, gadget A is what you use. Knowing
why you used that one, makes you come back and evaluate the
solution once an alternative is available. In other cases, one have
spent
years evaluating alternatives and optimizing solutions. Those
implementations a should probably be left alone.

> We became, if not buddies, respected acquaintances.

I am not surprised. You gave the non-bullshit answer. He recognized
a good engineer.

> I devised ways to
> water his flowers while away without limiting flow with orifices so
> small they would clog. I figured out how he could legally stop the
> Pitot-static airspeed indicator on his plane from freezing up without a
> mechanic's license. We brainstormed about problems I wanted to crack;
> usually about work, but not always.
>
> It seems that for all his bluster, what he wanted was someone who didn't
> flinch. (As a researcher, he had built the first power transistor -- one
> watt -- while a meeting was being held to discuss the feasibility of
> ever reaching a quarter watt. He had left the meeting impatiently and
> returned before it was over with a pellet soldered (by hand) to a copper
> plate. He chucked the thing still warm into the middle of the conference
> table and said, "One watt, ten minutes, still going strong when I
> disconnected it."

Yep, seems slike a guy that could intimidate his peers. I wouldn't be
surprised
if, say, three or four guys like you two would outpreform an entire R&D

department, if given decent resources.

What happened to this guy? Did he reach retirement? 

Rune

Rune Allnor wrote:
> Hi all.
> 
> Many years ago, I attended some conference where some guy who did
> not have English as his "natural 2nd language" gave a presentation.
> This guy was the only one I saw during a full week to actually keep
> the time schedule. He wrapped up his talk at exactly 15:00 minutes,
> leaving five full minutes for questions and discussions. No questions
> were asked from the audience, possibly because of the presenter's
> not quite perfect English diction. Not that I blame him for that, he
> did a better job than I would have managed if I was told to give a
> presentation in Russian, Chinese of Japanese in six months time.
> 
> anyway, the chairman, a slick elegant English academic of Oxbridge
> grooming, who was very aware he was a slick English academic of
> Oxbridge grooming, found the silence soewhat awkward, and asked,
> in a way only a slick English academic of Oxbridge grooming can,
> 
> "Excuse me sir, but what is new about this work?"
> 
> The reaction of the presenter was as if the chairman had pretended to
> shake his hand but hit him to the stomach instead. And we got a speedy
> almost verbatim re-run of the 15 minutes talk, this time in two
> minutes.
> 
> After the session I went up to the chairman and asked if he hadn't been
> 
> a bit rough with the poor guy. "Oh no," he said, "that is the *one*
> question
> everybody who gives a talk at a scientific conference should always
> be prepared for. If you can't answer it, you should not give the
> presentation in the first place."
> 
> It did make sense to me.
> 
> So I adapted it to my own version as "Excuse me sir, why do you think
> this
> method will work?" applied to engineering. True, the question is a bit
> tough
> if asked out loud, but I find it useful as a philosophical basis. If I
> can not
> formulate an answer to explain the intention of using some method or
> algorithm, I don't know why I use it and should probably not be doing
> the
> job in the first place.
> 
> That's the philosophical basis I have used for a decade. Whenever I
> have
> had a saying in a project, I have used the philosophy to suggest
> solutions
> when I can, and explain why I could not find any solution when that was
> 
> the case. Lots of people don't like to hear that something can *not* be
> 
> done. So I have more than once pointed out that Niels Henrik Abel,
> the only Norwegian mathematician of international fame, is most widely
> known in the general public, for having proved that it is *im*possible
> to
> solve fifth degree equations by means of the "standard" algebraic
> methods.
> I found some of the reactions to be all out disturbing.
> 
> It follows as a trivial corrolary that one should stay away from
> projects
> where no path towards a solution exists, unless the objective of the
> project is to find such a path. Basically, it is a matter of getting a
> pre-project
> feasibility study, which is part of basic project managment
> craftmanship.
> 
> Imagine my surprise, even digust, when I discussed some projects with
> som guy a few days ago. This guy said he was not the least interested
> in
> feasability studies. "They always conclude with that whatever can not
> be done." Which is just about true, what underwater acoustics is
> concerned.
> So no one write feasability studies. And no one validate the methods
> they come up with. Neither is good for business.
> 
> And this seems to be intentional. I cited one particular project, that
> was
> based on the following argument:
> 
> Use an active sonar to insonify the water. If target A is present, the
> recieved echo will show property X, or
> 
> A => X     [implication]
> 
> in logical formalism. I do not contest the implication.
> 
> The project was based on analyzing the recieved signal for property X,
> using this as an indicator for the target A. In doing this, one
> implicitly
> puts a constraint one can not defend:
> 
> A <=> X   [equivalence]
> 
> The equivalence means that only target A, and no others, induces
> property X in the recieved data. Which is a very strong claim.
> What needs to be done is to make sure there exist no other targets
> in the sea that show property X. Since "property X" in this particular
> case
> is formulated as "the target reflects frequencies f1,f2,...,fN" there
> is no
> way of establishing the equivalence. It is a ridiculous claim to say
> that
> one and only one type of object reflects a set of specified
> frequencies.
> 
> So I pointed this out. There are most probably several targets that
> show property X,
> 
> A => X
> B => X
> C => X
> 
> and so on ad infinitum. So making an observation X can not be
> linked to target A in any definite way, basically kicking the feet
> away beneath the project in question. "Yep, that's logical" was
> the answer. Sure, it is. Students are required to take a basic course
> of elementary logic before being allowed to take any degree in
> a university. But the connection to the real-world project in question
> was not grasped.
> 
> I just don't get it. Somebody proposes and sells a project where
> there is absolutely no chance of success, whatsoever. When asked,
> they see the argument why, and agree with it, but still go on. Why?
> Do they have no professional pride at all? Is it worth the carreer
> and reputation of the junior engineers that are assigned to do the
> grunt work? There is a peculiar pattern in these types of projects,
> that the basic ideas behind the projects are never criticized when
> they blow up. It is always the assigned engineer who is not up to
> the task.
> 
> Oh well.
> 
> There must be material for at least a couple of doctoral theses in
> psychology, somewhere inside here.
> 
> Rune
> 
I read a book a number of years called "Conceptual Block Busting" which 
is supposed to help you learn how to be "creative".  I actually liked 
the book.  One thing it suggests is to not throw away ideas too fast.

A lot of organizations have more of less institutionalized the concept. 
  There is no doubt in my mind that some people cynically exploit it, 
but  that doesn't mean that the process is totally dysfunctional.

Rune Allnor wrote:
> Jerry Avins wrote:

   ...

>>Then one ought to say that one should stay away from projects where no
>>path towards a solution is known to exist.
> 
> 
> If the project has committed to deliver a working solution, then yes.
> Most projects I have been involved with blew up because such
> commitments were made, and no solutions could be found.

Committing to deliver a thing with no precedent in a specified time at a 
specified cost is no more ethical than a guaranteeing the return on a 
speculative investment of money.

   ...

>>                  I worked in a research organization. I had to
>>tell a new boss that it was perfectly all right to assign me a job with
>>clear goals, tell me how it would be accomplished, and how long it will
>>take. He had to realize, though, that the last two specifications were
>>mere predictions. Moreover, Whether they proved accurate or not, the
>>project could hardly be called "research".
> 
> 
> I don't agree with that. I agree with being assigned a goal and a
> time/cost horizon.

That depends on how speculative the project is. It's easy to set goals 
that can't be met, or which, in some important sense, are already met. 
Interesting assignments fall between those extremes.

>                   I do not, from consistent experience over many
> years, trust anybody who tells me how to do things. If somebody
> wants my cooperation, tell me what resources I have available,
> and leave the rest to me.
 >
> Including time and resources for a feasability study to discover
> if, and if so how, the goals can be achieved within the constraints.

Such a study implies that the goal might not, after all, be met. For 
some rah-rah types, that's not an acceptable way to think.

   ...
>>     If you were bullshitting, he would know before anyone else,
>>maybe even yourself.
> 
> 
> What do you mean by "bullshiting"? Pretending you had answers you
> did not? Doing useless work?

Presenting oneself in more flattering terms than warranted; claiming 
forethought when there was none; exaggerating the importance of trivial 
things.

> That's the key point of my starting anecdote: SOME choise have to be
> made. If gadget A is what you have, gadget A is what you use. Knowing
> why you used that one, makes you come back and evaluate the solution
> once an alternative is available. In other cases, one have spent
> years evaluating alternatives and optimizing solutions. Those
> implementations a should probably be left alone.
> 
> 
>>We became, if not buddies, respected acquaintances.
> 
> 
> I am not surprised. You gave the non-bullshit answer. He recognized
> a good engineer.

   ...

> Yep, seems slike a guy that could intimidate his peers. I wouldn't be
> surprised
> if, say, three or four guys like you two would outpreform an entire R&D
> 
> department, if given decent resources.
> 
> What happened to this guy? Did he reach retirement? 

Yes, but not in good health. Too bad.

Jerry
-- 
Engineering is the art of making what you want from things you can get.
�����������������������������������������������������������������������

Peter K. wrote:
> "Rune Allnor" <allnor@tele.ntnu.no> writes:
>
> > I just don't get it. Somebody proposes and sells a project where
> > there is absolutely no chance of success, whatsoever. When asked,
> > they see the argument why, and agree with it, but still go on. Why?
> > Do they have no professional pride at all? Is it worth the carreer
> > and reputation of the junior engineers that are assigned to do the
> > grunt work? There is a peculiar pattern in these types of projects,
> > that the basic ideas behind the projects are never criticized when
> > they blow up. It is always the assigned engineer who is not up to
> > the task.
>
> Several times I've seen problems which, when interpreted one way, have
> no solution.  However, the statement of the problem is usually
> ambiguous enough to allow for other interpretations.  This allows
> "cheating" (re-scoping of the problem so that it is tractable).
>
> Cheating in this way can be good fruit for research or innovation. :-)

It depends on how the project description is phrased. If somebody
have a problem (i.e. the user requests help) and says "this is what
we have and that is what we want" one can usually find some
way of improving the user's situation. Maybe not quite to the
ideal situation one would really want (a sonar that classifies fish
in terms of species and size is all but improssible to make), but
something that is better than what is used today (maybe something
that tracks schools or individuals of fish better than one can today).

The problem occurs when somebody presents a solution to a
problem on a flawed basis, and does not allow for errors in
the problem formulation. "Wild and farmed salmon (Salmo Salar)
are different types (note the word "type", not "species") of fish,
and need to be separated to administer salmon stocks properly.
We will make a sonar that classifies salmon as either wild or
farmed based on the acoustic responses." If somebody thinks
this ought to be possible, check out an image of wild and farmed
salmon (salmo salar) like this one:

http://www.norwegian-salmon.com/salmon/extended.php?recID=3D87

To make it really interesting, compare the two salmons to the sea
trout (salmo trutta), here

http://birkenes.org/index.php?value=3Dfisker

The Norwegian name is "sj=F8=F8rret", the image is number 7 from
the top. The trout is a completely different species than salmon,
and has nothing to do with the wild/farmed salmon debate.
But it looks virtually identical.

I haven't seen the species salmon discussed in any project (yet),
but otherwise the above project description is completely in line
with ongoing projects I know of.

Rune

Stan Pawlukiewicz wrote:
> I read a book a number of years called "Conceptual Block Busting" which
> is supposed to help you learn how to be "creative".  I actually liked
> the book.  One thing it suggests is to not throw away ideas too fast.
>
> A lot of organizations have more of less institutionalized the concept.
>   There is no doubt in my mind that some people cynically exploit it,
> but  that doesn't mean that the process is totally dysfunctional.

Hmmmm... I don't know the book, but I don't believe creativity
can be learned on the individual level. Either you have some sort
of creative talent, or you don't. Of course, "creative talent" is an
ambiguous term. A Garry Kasparov can find a "creative" attac
when he plays a game of chess, and a Jimi Hendrix could find
"creative" new ways of handling a guitar. I don't think either
form of creativity can be taught to somebody who does not
already have some talent along the same lines.

It might be useful in organizations to learn to recognize creative
staff, and let them use their creativity, though. But that does not
mean that all "creative" ideas are good ideas.

Rune