DNA animation

[whit3rd]

On Wednesday, May 15, 2019 at 12:14:55 PM UTC-7, John Larkin wrote:

On Tue, 14 May 2019 20:25:29 -0700 (PDT), whit3rd <whit3rd@gmail.com
wrote:

On Tuesday, May 14, 2019 at 4:32:47 PM UTC-7, John Larkin wrote:

Because science keeps being blindsided by astounding discoveries.

Huh? Science is NOT blind to possibilities,

Sadly, sometimes not just blind but outright hostile.

Nonsense; hostility is personal, science is not.
You're thinking about 'stuffed-shirt professor stereotype'
and not science.

those discoveries are
the result of planning and careful work.

Or some crazy amateur who doesn't know he/she isn't allowed to
discover things.

Amateur means the person enjoys the task and doesn't seek other reward,
not relevant to 'being a scientist'. You're thinking about 'heroic/manic innovator stereotype'
not amateur. It makes a good story when the innovator makes his discovery,
but it's only a storytelling trope, not common in practice.

Stanford Ovshinsky's brown solar cells were a real discovery, and science welcomed
it only after some opening of the 'trade secrets'. There's not a lot of other examples
that come to mind (the Starlite heat shield material is still a mystery).

 

[John Larkin]

On Thu, 16 May 2019 10:38:15 +0100, Martin Brown
<'''newspam'''@nezumi.demon.co.uk> wrote:

On 15/05/2019 15:41, John Larkin wrote:
On Wed, 15 May 2019 07:18:16 -0700 (PDT), George Herold
gherold@teachspin.com> wrote:

On Wednesday, May 15, 2019 at 9:36:56 AM UTC-4, Martin Brown wrote:

The most interesting one would be to see if we can generate the right
structures in a human to permit chloroplasts and photosynthesis. We
wouldn't need to eat quite so much if we could directly make sugars.

And green humans like the Treens in Dan Dare would be quite cool.

Without doing any numbers.. it seems like there would hardly be any gain.
(I'm lucky to get ~10 hours of full sun in a week.)
It takes a corn plant all summer to make a few ears of corn.
(I'm guessing I have about the same area as a corn plant.)

And not wear clothes...

Get as much sun as you can. It not only makes vitamin D, it does other
good stuff.

Some sun is good but even in the UK which is a fairly high latitude too
much of it can prematurely age the skin and cause malignant skin cancer
if you are unlucky. Rickets is making a come back thanks to very high
factor sun protection being used on children these days.

Friends with very fair skins turn red very quickly if out in the sun.

MS is unheard of in sunny climes. People in cold gloomy places get it.

There certainly does seem to be a latitude correlation but there are
other risk factors like genetic susceptibility and possibly infection by
the Epstein-Barr virus. Being female seems to carry a serious risk too.

https://www.nhs.uk/conditions/multiple-sclerosis/causes/

Yes. My wife grew up in Boston, got MS after we were married, and
miraculously grew out of it. Sometimes that happens.


--

John Larkin Highland Technology, Inc

lunatic fringe electronics

 

[John Larkin]

On Thu, 16 May 2019 07:38:40 +0100, Tom Gardner
<spamjunk@blueyonder.co.uk> wrote:

On 16/05/19 03:43, Bill Sloman wrote:
On Thursday, May 16, 2019 at 2:06:59 AM UTC+10, John Larkin wrote:
On Wed, 15 May 2019 15:54:01 +0100, Tom Gardner <spamjunk@blueyonder.co.uk
wrote:

On 15/05/19 15:32, John Larkin wrote:
I don't think we'll invite you to any of our brainstorming sessions.
Some people poison brainstorming.

Yebbut. There are two phases to brainstorming: - firstly rapid generation
of ideas, which requires complete suspension of disbelief - followed by
selection of the ideas that might work, and discarding the others

Alternatively consider team makeup...

If you have two "ideas men" only, then sparks will fly and everybody will
also have great fun - but nothing will be able to come of it.

If you have two "critics" only, then there will be very realistic plans,
but they will be boring.

Sometimes the seed of a great idea comes from someone that nobody expected
anything from, like an intern invited in to observe. Sometimes the
inspiration is just a question.

Example?

I forget all the details, but once upon a time we
were brainstorming comms systems, and things were
in the process of stalling.

I asked "how would you do that with yoghurt", and
things got moving again - and reached a useful conclusion.

Trivial? Yes, of course.

Feynman told the story of being sent out to assess
a production facility for making the bomb in WW2.
He knew he wouldn't be able to provide any detailed
technical assessment. He did randomly select a valve
and ask what would happen if it jammed. That kicked
off a discussion amongst the local staff, and they
did discover a significant vulnerability.



Brainstorming is a group extension of a basic process: send your mental
tendrils as far and wide as possible into the potential, real or absurd,
solution space, dredge up anything interesting or amusing, and play with it
to see what develops. More people can spread out further into that space,
or riff on what someone else finds.

Some people extend into more useful space than others. Knowing more about
what you are brainstorming about can point the extension into more useful
areas.

Yes, but the ability of someone "with a different
toolkit" or who isn't "in the middle of the trees"
can be significantly helpful.

I'm sure you can think of such cases from your own
experience.

I naturally enjoy the ones where one or two elecronics people (ie, me
and maybe my allies) are presented with a problem owned by
non-electronics people, like chemists or physicists or MEs.

AoE should be mandatory study for any scientist. Everything is
electronic these days, and it's hard to imagine (irony) anything
replacing the universality of electronic instrumentation in science.


--

John Larkin Highland Technology, Inc

lunatic fringe electronics

 

[Tom Gardner]

On 16/05/19 15:13, John Larkin wrote:

On Thu, 16 May 2019 07:38:40 +0100, Tom Gardner
spamjunk@blueyonder.co.uk> wrote:

On 16/05/19 03:43, Bill Sloman wrote:
On Thursday, May 16, 2019 at 2:06:59 AM UTC+10, John Larkin wrote:
On Wed, 15 May 2019 15:54:01 +0100, Tom Gardner <spamjunk@blueyonder.co.uk
wrote:

On 15/05/19 15:32, John Larkin wrote:
I don't think we'll invite you to any of our brainstorming sessions.
Some people poison brainstorming.

Yebbut. There are two phases to brainstorming: - firstly rapid generation
of ideas, which requires complete suspension of disbelief - followed by
selection of the ideas that might work, and discarding the others

Alternatively consider team makeup...

If you have two "ideas men" only, then sparks will fly and everybody will
also have great fun - but nothing will be able to come of it.

If you have two "critics" only, then there will be very realistic plans,
but they will be boring.

Sometimes the seed of a great idea comes from someone that nobody expected
anything from, like an intern invited in to observe. Sometimes the
inspiration is just a question.

Example?

I forget all the details, but once upon a time we
were brainstorming comms systems, and things were
in the process of stalling.

I asked "how would you do that with yoghurt", and
things got moving again - and reached a useful conclusion.

Trivial? Yes, of course.

Feynman told the story of being sent out to assess
a production facility for making the bomb in WW2.
He knew he wouldn't be able to provide any detailed
technical assessment. He did randomly select a valve
and ask what would happen if it jammed. That kicked
off a discussion amongst the local staff, and they
did discover a significant vulnerability.



Brainstorming is a group extension of a basic process: send your mental
tendrils as far and wide as possible into the potential, real or absurd,
solution space, dredge up anything interesting or amusing, and play with it
to see what develops. More people can spread out further into that space,
or riff on what someone else finds.

Some people extend into more useful space than others. Knowing more about
what you are brainstorming about can point the extension into more useful
areas.

Yes, but the ability of someone "with a different
toolkit" or who isn't "in the middle of the trees"
can be significantly helpful.

I'm sure you can think of such cases from your own
experience.

I naturally enjoy the ones where one or two elecronics people (ie, me
and maybe my allies) are presented with a problem owned by
non-electronics people, like chemists or physicists or MEs.

AoE should be mandatory study for any scientist. Everything is
electronic these days, and it's hard to imagine (irony) anything
replacing the universality of electronic instrumentation in science.

Half a lifetime ago (early 80s) I did a consultancy gig to
look at replacing logic with a microprocessor etc. I
recommended they did /not/ replace it.

The logic was on unmanned offshore oil rigs, with no
electricity on board. Electronics would have introduced
BASEEFA(?) requirements, and had no advantages over the
existing /hydraulic/ logic.

Yes, I did mention the advantages of remote monitoring/control,
but the internet was in its infancy then, e.g. email used bang
addressing.

 

[Rick C]

On Thursday, May 16, 2019 at 10:13:13 AM UTC-4, John Larkin wrote:

On Thu, 16 May 2019 07:38:40 +0100, Tom Gardner
spamjunk@blueyonder.co.uk> wrote:

On 16/05/19 03:43, Bill Sloman wrote:
On Thursday, May 16, 2019 at 2:06:59 AM UTC+10, John Larkin wrote:
On Wed, 15 May 2019 15:54:01 +0100, Tom Gardner <spamjunk@blueyonder.co.uk
wrote:

On 15/05/19 15:32, John Larkin wrote:
I don't think we'll invite you to any of our brainstorming sessions.
Some people poison brainstorming.

Yebbut. There are two phases to brainstorming: - firstly rapid generation
of ideas, which requires complete suspension of disbelief - followed by
selection of the ideas that might work, and discarding the others

Alternatively consider team makeup...

If you have two "ideas men" only, then sparks will fly and everybody will
also have great fun - but nothing will be able to come of it.

If you have two "critics" only, then there will be very realistic plans,
but they will be boring.

Sometimes the seed of a great idea comes from someone that nobody expected
anything from, like an intern invited in to observe. Sometimes the
inspiration is just a question.

Example?

I forget all the details, but once upon a time we
were brainstorming comms systems, and things were
in the process of stalling.

I asked "how would you do that with yoghurt", and
things got moving again - and reached a useful conclusion.

Trivial? Yes, of course.

Feynman told the story of being sent out to assess
a production facility for making the bomb in WW2.
He knew he wouldn't be able to provide any detailed
technical assessment. He did randomly select a valve
and ask what would happen if it jammed. That kicked
off a discussion amongst the local staff, and they
did discover a significant vulnerability.



Brainstorming is a group extension of a basic process: send your mental
tendrils as far and wide as possible into the potential, real or absurd,
solution space, dredge up anything interesting or amusing, and play with it
to see what develops. More people can spread out further into that space,
or riff on what someone else finds.

Some people extend into more useful space than others. Knowing more about
what you are brainstorming about can point the extension into more useful
areas.

Yes, but the ability of someone "with a different
toolkit" or who isn't "in the middle of the trees"
can be significantly helpful.

I'm sure you can think of such cases from your own
experience.

I naturally enjoy the ones where one or two elecronics people (ie, me
and maybe my allies) are presented with a problem owned by
non-electronics people, like chemists or physicists or MEs.

AoE should be mandatory study for any scientist. Everything is
electronic these days, and it's hard to imagine (irony) anything
replacing the universality of electronic instrumentation in science.

Like most skills and professions, they can always hire someone to do what is needed in electronics. There is little magic left. Most things can be done by buying modules.

--

Rick C.

+--+- Get 5,000 miles of free Supercharging
+--+- Tesla referral code - https://ts.la/richard11209

 

[John Larkin]

On Thu, 16 May 2019 16:29:40 +0100, Tom Gardner
<spamjunk@blueyonder.co.uk> wrote:

On 16/05/19 15:13, John Larkin wrote:
On Thu, 16 May 2019 07:38:40 +0100, Tom Gardner
spamjunk@blueyonder.co.uk> wrote:

On 16/05/19 03:43, Bill Sloman wrote:
On Thursday, May 16, 2019 at 2:06:59 AM UTC+10, John Larkin wrote:
On Wed, 15 May 2019 15:54:01 +0100, Tom Gardner <spamjunk@blueyonder.co.uk
wrote:

On 15/05/19 15:32, John Larkin wrote:
I don't think we'll invite you to any of our brainstorming sessions.
Some people poison brainstorming.

Yebbut. There are two phases to brainstorming: - firstly rapid generation
of ideas, which requires complete suspension of disbelief - followed by
selection of the ideas that might work, and discarding the others

Alternatively consider team makeup...

If you have two "ideas men" only, then sparks will fly and everybody will
also have great fun - but nothing will be able to come of it.

If you have two "critics" only, then there will be very realistic plans,
but they will be boring.

Sometimes the seed of a great idea comes from someone that nobody expected
anything from, like an intern invited in to observe. Sometimes the
inspiration is just a question.

Example?

I forget all the details, but once upon a time we
were brainstorming comms systems, and things were
in the process of stalling.

I asked "how would you do that with yoghurt", and
things got moving again - and reached a useful conclusion.

Trivial? Yes, of course.

Feynman told the story of being sent out to assess
a production facility for making the bomb in WW2.
He knew he wouldn't be able to provide any detailed
technical assessment. He did randomly select a valve
and ask what would happen if it jammed. That kicked
off a discussion amongst the local staff, and they
did discover a significant vulnerability.



Brainstorming is a group extension of a basic process: send your mental
tendrils as far and wide as possible into the potential, real or absurd,
solution space, dredge up anything interesting or amusing, and play with it
to see what develops. More people can spread out further into that space,
or riff on what someone else finds.

Some people extend into more useful space than others. Knowing more about
what you are brainstorming about can point the extension into more useful
areas.

Yes, but the ability of someone "with a different
toolkit" or who isn't "in the middle of the trees"
can be significantly helpful.

I'm sure you can think of such cases from your own
experience.

I naturally enjoy the ones where one or two elecronics people (ie, me
and maybe my allies) are presented with a problem owned by
non-electronics people, like chemists or physicists or MEs.

AoE should be mandatory study for any scientist. Everything is
electronic these days, and it's hard to imagine (irony) anything
replacing the universality of electronic instrumentation in science.

Half a lifetime ago (early 80s) I did a consultancy gig to
look at replacing logic with a microprocessor etc. I
recommended they did /not/ replace it.

The logic was on unmanned offshore oil rigs, with no
electricity on board. Electronics would have introduced
BASEEFA(?) requirements, and had no advantages over the
existing /hydraulic/ logic.

Yes, I did mention the advantages of remote monitoring/control,
but the internet was in its infancy then, e.g. email used bang
addressing.

I designed some really slick electronic boiler and throttle controls
for the LHA ships. The crew couldn't maintain them, so the navy ripped
it all out and installed pneumatics.


--

John Larkin Highland Technology, Inc

lunatic fringe electronics

 

On Thu, 16 May 2019 08:58:34 +0100, Martin Brown
<'''newspam'''@nezumi.demon.co.uk> wrote:

On 16/05/2019 02:21, krw@notreal.com wrote:
On Wed, 15 May 2019 11:02:34 +0100, Martin Brown
'''newspam'''@nezumi.demon.co.uk> wrote:

On 15/05/2019 00:36, krw@notreal.com wrote:
On Tue, 14 May 2019 21:45:36 +0100, Tom Gardner
spamjunk@blueyonder.co.uk> wrote:

On 14/05/19 17:00, John Larkin wrote:
Of course it's a big problem. Big problems need big ideas.

Big /solutions/ need falsifiable hypotheses and tests.

Exactly what problem are you trying to solve, with the origin of life?

How did it get started? Where should be be looking for other life?

How is that a "problem" that _needs_ to be "solved"?

Nothing *needs* to be solved, but scientists are always curious about
how the universe works we could all worship at your Tree of Ignorance
and be significantly worse off as a result.

If you'd said they wanted an answer, rather than a problem solved, it
would have made much more sense. It's your ignorance we're discussing
here, BTW.

You never know where blue skies research will leads in the long term.

What earthly use could a laser be if the first one required a huge flash
tube and a 4" perfect ruby crystal to make it work ever be to anybody?

Now they are ubiquitous and in every laser printer, CD and DVD player.

4" perfect ruby crystals?

 

On Thu, 16 May 2019 14:03:12 +1000, Clifford Heath
<no.spam@please.net> wrote:

On 16/5/19 1:08 pm, Bill Sloman wrote:
On Thursday, May 16, 2019 at 11:22:05 AM UTC+10, k...@notreal.com wrote:
On Wed, 15 May 2019 11:02:34 +0100, Martin Brown
'''newspam'''@nezumi.demon.co.uk> wrote:

On 15/05/2019 00:36, krw@notreal.com wrote:
On Tue, 14 May 2019 21:45:36 +0100, Tom Gardner
spamjunk@blueyonder.co.uk> wrote:

On 14/05/19 17:00, John Larkin wrote:

snip

There is an optimum size of difficulty of problems that can be tackled
with today's available resources. So long as Moore's Law holds you can
prove that for some hard computational problems the fastest way to the
solution is to go surfing on the beach for a couple of years and then
start building your hardware using the latest fastest CPUs and memory.

That's like saying "I'll be later getting to work if I leave now,
rather than an hour from now because the traffic is worse this time of
the day.". Absurd.

No, it's like saying "should I jump on the bus now, or wait until my
wife returns from the shops so I can drive the car to work?"'

Bullshit

If you start work with a technology that slows you down, it's easy to
get too invested in it to be able to jump ship when a better way comes
along - so it's better not to start yet.

It doesn't "slow you down". There is nothing saying that you can't
change horses. Absurd.

Wrong analogy. I've jumped in early twice - once using TI's 64k serial memory (in 1978) and once using GigaBit Logic's GaAs fast logic (in 1988).

In the first case, 16k RAM got cheap enough a few months later to make the approach sub-optimal, and in the second case GaAs logic never got up to the production yields it needed to make it attractive while Motorola's ECLinPS was close enough behind (and much easier to produce - and use) to kill any enthusiasm for further development.

It's always hard to say though. At the time when Intel produced the
first Pentiums and had trouble getting the yields high enough, they were
using 17 mask layers to make it.

DEC's Alpha was using 4, HP's 800 series was using 3, and MIPS was using
only 2, to produce CPUs of comparable power - but with much better
yields of course, and much better MIPS/Watt.

Where are those others, now?

And before that, we thought it was all about MIPS, until we discovered
that it was actually all about bandwidth - the RISC vs CISC wars died
out when neither could get data on and off chip as fast as they could do
something useful with it.

Clifford Heath.

 

[Bill Sloman]

On Friday, May 17, 2019 at 11:33:34 AM UTC+10, k...@notreal.com wrote:

On Thu, 16 May 2019 08:58:34 +0100, Martin Brown
'''newspam'''@nezumi.demon.co.uk> wrote:

On 16/05/2019 02:21, krw@notreal.com wrote:
On Wed, 15 May 2019 11:02:34 +0100, Martin Brown
'''newspam'''@nezumi.demon.co.uk> wrote:

On 15/05/2019 00:36, krw@notreal.com wrote:
On Tue, 14 May 2019 21:45:36 +0100, Tom Gardner
spamjunk@blueyonder.co.uk> wrote:

On 14/05/19 17:00, John Larkin wrote:
Of course it's a big problem. Big problems need big ideas.

Big /solutions/ need falsifiable hypotheses and tests.

Exactly what problem are you trying to solve, with the origin of life?

How did it get started? Where should be be looking for other life?

How is that a "problem" that _needs_ to be "solved"?

Nothing *needs* to be solved, but scientists are always curious about
how the universe works we could all worship at your Tree of Ignorance
and be significantly worse off as a result.

If you'd said they wanted an answer, rather than a problem solved, it
would have made much more sense.

You've got to recognise that there's a problem to be solved before you can start looking for an answer. The most interesting phrase for a scientist is "that's odd ... ". What makes "sense" to krw is stuff that fits his totally pre-programmed brain, where every question has it's built-in answer.

The real world is a little more complicated, but he's not equipped to engage with that.

It's your ignorance we're discussing
here, BTW.

Actually, John Larkin's ignorance. His brain is better than krw's in that it can recognise that unsolved problems exist, but it's capacity to distinguish between plausible answers and self-serving propaganda seems to be non-existent.

You never know where blue skies research will leads in the long term.

What earthly use could a laser be if the first one required a huge flash
tube and a 4" perfect ruby crystal to make it work ever be to anybody?

Now they are ubiquitous and in every laser printer, CD and DVD player.

4" perfect ruby crystals?

That particular kind of laser isn't used laser printers, compact disk players or digital video disk players. People with more mental depth than krw are aware of this.

--
Bill Sloman, Sydney

 

[Bill Sloman]

On Friday, May 17, 2019 at 11:31:38 AM UTC+10, k...@notreal.com wrote:

On Thu, 16 May 2019 14:03:12 +1000, Clifford Heath
no.spam@please.net> wrote:

On 16/5/19 1:08 pm, Bill Sloman wrote:
On Thursday, May 16, 2019 at 11:22:05 AM UTC+10, k...@notreal.com wrote:
On Wed, 15 May 2019 11:02:34 +0100, Martin Brown
'''newspam'''@nezumi.demon.co.uk> wrote:

On 15/05/2019 00:36, krw@notreal.com wrote:
On Tue, 14 May 2019 21:45:36 +0100, Tom Gardner
spamjunk@blueyonder.co.uk> wrote:

On 14/05/19 17:00, John Larkin wrote:

snip

There is an optimum size of difficulty of problems that can be tackled
with today's available resources. So long as Moore's Law holds you can
prove that for some hard computational problems the fastest way to the
solution is to go surfing on the beach for a couple of years and then
start building your hardware using the latest fastest CPUs and memory.

That's like saying "I'll be later getting to work if I leave now,
rather than an hour from now because the traffic is worse this time of
the day.". Absurd.

No, it's like saying "should I jump on the bus now, or wait until my
wife returns from the shops so I can drive the car to work?"'

Bullshit

Which means that krw can't follow the logic, which isn't surprising - the logic isn't demanding but krw's brain doesn't seem to work.

If you start work with a technology that slows you down, it's easy to
get too invested in it to be able to jump ship when a better way comes
along - so it's better not to start yet.

It doesn't "slow you down". There is nothing saying that you can't
change horses. Absurd.

Nobody says that you can't change horses, but you do have to pay a bundle for the extensive redesign required to accommodate new and better technology.. It's not as expensive as designing from scratch, but it's still quite expensive enough to sink most projects.

<snipped the stuff that krw didn't read about my experiences where I couldn't "change horses".>

--
Bill Sloman, Sydney

 

[Martin Brown]

On 17/05/2019 02:31, krw@notreal.com wrote:

On Thu, 16 May 2019 14:03:12 +1000, Clifford Heath
no.spam@please.net> wrote:

On 16/5/19 1:08 pm, Bill Sloman wrote:
On Thursday, May 16, 2019 at 11:22:05 AM UTC+10, k...@notreal.com wrote:
On Wed, 15 May 2019 11:02:34 +0100, Martin Brown
'''newspam'''@nezumi.demon.co.uk> wrote:

On 15/05/2019 00:36, krw@notreal.com wrote:
On Tue, 14 May 2019 21:45:36 +0100, Tom Gardner
spamjunk@blueyonder.co.uk> wrote:

On 14/05/19 17:00, John Larkin wrote:

snip

There is an optimum size of difficulty of problems that can be tackled
with today's available resources. So long as Moore's Law holds you can
prove that for some hard computational problems the fastest way to the
solution is to go surfing on the beach for a couple of years and then
start building your hardware using the latest fastest CPUs and memory.

That's like saying "I'll be later getting to work if I leave now,
rather than an hour from now because the traffic is worse this time of
the day.". Absurd.

No, it's like saying "should I jump on the bus now, or wait until my
wife returns from the shops so I can drive the car to work?"'

Bullshit

You really are dumber than a rock!

Here is a concrete example taken from the time when the typical
improvement in clock speed and performance was roughly a 3 fold increase
for every 5 years elapsed. Once you start running the software you are
then committed to the technology that you begin to work with.

Runtime for computation A = 10 years.
Initial performance p = 1
Annual performance improvement factor a = 3^(1/5)
Wait time before starting execution t = 0

Total time T = t + A/(a^t)

Start computing immediately - wait time t = 0

Total time T = 0 + 10/1 = 10

Wait 5 years t = 5

Total time T = 5 + 10/3 = 8.333'

In fact for a device improving with an exponential law this can be
generalised to find the optimum time to start to finish soonest.

T = t + A/(a^t)

dT/dt = 0 = 1 - Aln(a)/a^t

Hence t = ln(Aln(a))/ln(a) = 3.5827 when T = 8.1339

So you can get your difficult result nearly 2 years earlier by spending
the first three and a half years doing something else more fun until the
computing hardware available is up to the job of running your code.

--
Regards,
Martin Brown

 

[Andy Bennet]

On 08/05/2019 05:24, John Larkin wrote:

https://www.youtube.com/watch?v=fpHaxzroYxg

This is insane. This is impossible.

I'll stick to my D-Types and Nand gates thanks!

 

On Fri, 17 May 2019 08:41:50 +0100, Martin Brown
<'''newspam'''@nezumi.demon.co.uk> wrote:

On 17/05/2019 02:31, krw@notreal.com wrote:
On Thu, 16 May 2019 14:03:12 +1000, Clifford Heath
no.spam@please.net> wrote:

On 16/5/19 1:08 pm, Bill Sloman wrote:
On Thursday, May 16, 2019 at 11:22:05 AM UTC+10, k...@notreal.com wrote:
On Wed, 15 May 2019 11:02:34 +0100, Martin Brown
'''newspam'''@nezumi.demon.co.uk> wrote:

On 15/05/2019 00:36, krw@notreal.com wrote:
On Tue, 14 May 2019 21:45:36 +0100, Tom Gardner
spamjunk@blueyonder.co.uk> wrote:

On 14/05/19 17:00, John Larkin wrote:

snip

There is an optimum size of difficulty of problems that can be tackled
with today's available resources. So long as Moore's Law holds you can
prove that for some hard computational problems the fastest way to the
solution is to go surfing on the beach for a couple of years and then
start building your hardware using the latest fastest CPUs and memory.

That's like saying "I'll be later getting to work if I leave now,
rather than an hour from now because the traffic is worse this time of
the day.". Absurd.

No, it's like saying "should I jump on the bus now, or wait until my
wife returns from the shops so I can drive the car to work?"'

Bullshit

You really are dumber than a rock!

You're full of shit. But we all knew that.

Here is a concrete example taken from the time when the typical
improvement in clock speed and performance was roughly a 3 fold increase
for every 5 years elapsed. Once you start running the software you are
then committed to the technology that you begin to work with.

Absolutely irrelevant.

Runtime for computation A = 10 years.
Initial performance p = 1
Annual performance improvement factor a = 3^(1/5)
Wait time before starting execution t = 0

Total time T = t + A/(a^t)

Start computing immediately - wait time t = 0

Total time T = 0 + 10/1 = 10

Wait 5 years t = 5

Total time T = 5 + 10/3 = 8.333'

In fact for a device improving with an exponential law this can be
generalised to find the optimum time to start to finish soonest.

T = t + A/(a^t)

dT/dt = 0 = 1 - Aln(a)/a^t

Hence t = ln(Aln(a))/ln(a) = 3.5827 when T = 8.1339

So you can get your difficult result nearly 2 years earlier by spending
the first three and a half years doing something else more fun until the
computing hardware available is up to the job of running your code.

All completely irrelevant. Stupid, in fact.

 

[Bill Sloman]

On Friday, May 17, 2019 at 11:33:34 AM UTC+10, k...@notreal.com wrote:

On Thu, 16 May 2019 08:58:34 +0100, Martin Brown
'''newspam'''@nezumi.demon.co.uk> wrote:

On 16/05/2019 02:21, krw@notreal.com wrote:
On Wed, 15 May 2019 11:02:34 +0100, Martin Brown
'''newspam'''@nezumi.demon.co.uk> wrote:

On 15/05/2019 00:36, krw@notreal.com wrote:
On Tue, 14 May 2019 21:45:36 +0100, Tom Gardner
spamjunk@blueyonder.co.uk> wrote:

On 14/05/19 17:00, John Larkin wrote:
Of course it's a big problem. Big problems need big ideas.

Big /solutions/ need falsifiable hypotheses and tests.

Exactly what problem are you trying to solve, with the origin of life?

How did it get started? Where should be be looking for other life?

How is that a "problem" that _needs_ to be "solved"?

Nothing *needs* to be solved, but scientists are always curious about
how the universe works we could all worship at your Tree of Ignorance
and be significantly worse off as a result.

If you'd said they wanted an answer, rather than a problem solved, it
would have made much more sense.

Only someone as dim and rigid as krw would see any difference between the two of expressing the same idea.

It's your ignorance we're discussing
here, BTW.

Krw's ignorance is profound, and he thinks that anybody who knows more than he does is ignorant of his gross over-simplifications. Comical.

You never know where blue skies research will leads in the long term.

What earthly use could a laser be if the first one required a huge flash
tube and a 4" perfect ruby crystal to make it work ever be to anybody?

Now they are ubiquitous and in every laser printer, CD and DVD player.

4" perfect ruby crystals?

https://en.wikipedia.org/wiki/Ruby_laser

Krw was around back then, but clearly not paying attention. I was a first year undergraduate, which didn't stop me prattling about lasers at time. If I'd known more, I'd have said less ...

--
Bill Sloman, Sydney

 

[Bill Sloman]

On Saturday, May 18, 2019 at 1:45:16 PM UTC+10, k...@notreal.com wrote:

On Fri, 17 May 2019 08:41:50 +0100, Martin Brown
'''newspam'''@nezumi.demon.co.uk> wrote:

On 17/05/2019 02:31, krw@notreal.com wrote:
On Thu, 16 May 2019 14:03:12 +1000, Clifford Heath
no.spam@please.net> wrote:

On 16/5/19 1:08 pm, Bill Sloman wrote:
On Thursday, May 16, 2019 at 11:22:05 AM UTC+10, k...@notreal.com wrote:
On Wed, 15 May 2019 11:02:34 +0100, Martin Brown
'''newspam'''@nezumi.demon.co.uk> wrote:

On 15/05/2019 00:36, krw@notreal.com wrote:
On Tue, 14 May 2019 21:45:36 +0100, Tom Gardner
spamjunk@blueyonder.co.uk> wrote:

On 14/05/19 17:00, John Larkin wrote:

snip

There is an optimum size of difficulty of problems that can be tackled
with today's available resources. So long as Moore's Law holds you can
prove that for some hard computational problems the fastest way to the
solution is to go surfing on the beach for a couple of years and then
start building your hardware using the latest fastest CPUs and memory.

That's like saying "I'll be later getting to work if I leave now,
rather than an hour from now because the traffic is worse this time of
the day.". Absurd.

No, it's like saying "should I jump on the bus now, or wait until my
wife returns from the shops so I can drive the car to work?"'

Bullshit

You really are dumber than a rock!

You're full of shit. But we all knew that.

Here is a concrete example taken from the time when the typical
improvement in clock speed and performance was roughly a 3 fold increase
for every 5 years elapsed. Once you start running the software you are
then committed to the technology that you begin to work with.

Absolutely irrelevant.

Which is to say that krw is too pig ignorant to see the point.

Runtime for computation A = 10 years.
Initial performance p = 1
Annual performance improvement factor a = 3^(1/5)
Wait time before starting execution t = 0

Total time T = t + A/(a^t)

Start computing immediately - wait time t = 0

Total time T = 0 + 10/1 = 10

Wait 5 years t = 5

Total time T = 5 + 10/3 = 8.333'

In fact for a device improving with an exponential law this can be
generalised to find the optimum time to start to finish soonest.

T = t + A/(a^t)

dT/dt = 0 = 1 - Aln(a)/a^t

Hence t = ln(Aln(a))/ln(a) = 3.5827 when T = 8.1339

So you can get your difficult result nearly 2 years earlier by spending
the first three and a half years doing something else more fun until the
computing hardware available is up to the job of running your code.

All completely irrelevant. Stupid, in fact.

Which is krw's way of saying that he doesn't understand it's relevance - and is best translated as an admission by krw that he is terminally stupid (though he's too stupid to realise quite how stupid he is.

--
Bill Sloman, Sydney

 

[Martin Brown]

On 18/05/2019 04:45, krw@notreal.com wrote:

On Fri, 17 May 2019 08:41:50 +0100, Martin Brown
'''newspam'''@nezumi.demon.co.uk> wrote:

On 17/05/2019 02:31, krw@notreal.com wrote:
On Thu, 16 May 2019 14:03:12 +1000, Clifford Heath
no.spam@please.net> wrote:

On 16/5/19 1:08 pm, Bill Sloman wrote:
On Thursday, May 16, 2019 at 11:22:05 AM UTC+10, k...@notreal.com wrote:
On Wed, 15 May 2019 11:02:34 +0100, Martin Brown
'''newspam'''@nezumi.demon.co.uk> wrote:

On 15/05/2019 00:36, krw@notreal.com wrote:
On Tue, 14 May 2019 21:45:36 +0100, Tom Gardner
spamjunk@blueyonder.co.uk> wrote:

On 14/05/19 17:00, John Larkin wrote:

snip

There is an optimum size of difficulty of problems that can be tackled
with today's available resources. So long as Moore's Law holds you can
prove that for some hard computational problems the fastest way to the
solution is to go surfing on the beach for a couple of years and then
start building your hardware using the latest fastest CPUs and memory.

That's like saying "I'll be later getting to work if I leave now,
rather than an hour from now because the traffic is worse this time of
the day.". Absurd.

No, it's like saying "should I jump on the bus now, or wait until my
wife returns from the shops so I can drive the car to work?"'

Bullshit

You really are dumber than a rock!

You're full of shit. But we all knew that.

And you are wilfully ignorant and stupid as well.

Here is a concrete example taken from the time when the typical
improvement in clock speed and performance was roughly a 3 fold increase
for every 5 years elapsed. Once you start running the software you are
then committed to the technology that you begin to work with.

Absolutely irrelevant.

You clearly fail to understand. That is not my problem!

Runtime for computation A = 10 years.
Initial performance p = 1
Annual performance improvement factor a = 3^(1/5)
Wait time before starting execution t = 0

Total time T = t + A/(a^t)

Start computing immediately - wait time t = 0

Total time T = 0 + 10/1 = 10

Wait 5 years t = 5

Total time T = 5 + 10/3 = 8.333'

In fact for a device improving with an exponential law this can be
generalised to find the optimum time to start to finish soonest.

T = t + A/(a^t)

dT/dt = 0 = 1 - Aln(a)/a^t

Hence t = ln(Aln(a))/ln(a) = 3.5827 when T = 8.1339

So you can get your difficult result nearly 2 years earlier by spending
the first three and a half years doing something else more fun until the
computing hardware available is up to the job of running your code.

All completely irrelevant. Stupid, in fact.

It applies to large scale projects like sequencing the human genome and
computing the mass difference of the proton and neutron ab initio from
QCD (which is the first time I recall seeing the issue discussed).

--
Regards,
Martin Brown

 

On Sun, 19 May 2019 16:55:35 +0100, Martin Brown
<'''newspam'''@nezumi.demon.co.uk> wrote:

On 18/05/2019 04:45, krw@notreal.com wrote:
On Fri, 17 May 2019 08:41:50 +0100, Martin Brown
'''newspam'''@nezumi.demon.co.uk> wrote:

On 17/05/2019 02:31, krw@notreal.com wrote:
On Thu, 16 May 2019 14:03:12 +1000, Clifford Heath
no.spam@please.net> wrote:

On 16/5/19 1:08 pm, Bill Sloman wrote:
On Thursday, May 16, 2019 at 11:22:05 AM UTC+10, k...@notreal.com wrote:
On Wed, 15 May 2019 11:02:34 +0100, Martin Brown
'''newspam'''@nezumi.demon.co.uk> wrote:

On 15/05/2019 00:36, krw@notreal.com wrote:
On Tue, 14 May 2019 21:45:36 +0100, Tom Gardner
spamjunk@blueyonder.co.uk> wrote:

On 14/05/19 17:00, John Larkin wrote:

snip

There is an optimum size of difficulty of problems that can be tackled
with today's available resources. So long as Moore's Law holds you can
prove that for some hard computational problems the fastest way to the
solution is to go surfing on the beach for a couple of years and then
start building your hardware using the latest fastest CPUs and memory.

That's like saying "I'll be later getting to work if I leave now,
rather than an hour from now because the traffic is worse this time of
the day.". Absurd.

No, it's like saying "should I jump on the bus now, or wait until my
wife returns from the shops so I can drive the car to work?"'

Bullshit

You really are dumber than a rock!

You're full of shit. But we all knew that.

And you are wilfully ignorant and stupid as well.

NO, you're simply illiterate.

Here is a concrete example taken from the time when the typical
improvement in clock speed and performance was roughly a 3 fold increase
for every 5 years elapsed. Once you start running the software you are
then committed to the technology that you begin to work with.

Absolutely irrelevant.

You clearly fail to understand. That is not my problem!

Not much at thinking, either.

Runtime for computation A = 10 years.
Initial performance p = 1
Annual performance improvement factor a = 3^(1/5)
Wait time before starting execution t = 0

Total time T = t + A/(a^t)

Start computing immediately - wait time t = 0

Total time T = 0 + 10/1 = 10

Wait 5 years t = 5

Total time T = 5 + 10/3 = 8.333'

In fact for a device improving with an exponential law this can be
generalised to find the optimum time to start to finish soonest.

T = t + A/(a^t)

dT/dt = 0 = 1 - Aln(a)/a^t

Hence t = ln(Aln(a))/ln(a) = 3.5827 when T = 8.1339

So you can get your difficult result nearly 2 years earlier by spending
the first three and a half years doing something else more fun until the
computing hardware available is up to the job of running your code.

All completely irrelevant. Stupid, in fact.

It applies to large scale projects like sequencing the human genome and
computing the mass difference of the proton and neutron ab initio from
QCD (which is the first time I recall seeing the issue discussed).

 

[Bill Sloman]

On Monday, May 20, 2019 at 11:44:08 AM UTC+10, k...@notreal.com wrote:

On Sun, 19 May 2019 16:55:35 +0100, Martin Brown
'''newspam'''@nezumi.demon.co.uk> wrote:

On 18/05/2019 04:45, krw@notreal.com wrote:
On Fri, 17 May 2019 08:41:50 +0100, Martin Brown
'''newspam'''@nezumi.demon.co.uk> wrote:

On 17/05/2019 02:31, krw@notreal.com wrote:
On Thu, 16 May 2019 14:03:12 +1000, Clifford Heath
no.spam@please.net> wrote:

On 16/5/19 1:08 pm, Bill Sloman wrote:
On Thursday, May 16, 2019 at 11:22:05 AM UTC+10, k...@notreal.com wrote:
On Wed, 15 May 2019 11:02:34 +0100, Martin Brown
'''newspam'''@nezumi.demon.co.uk> wrote:

On 15/05/2019 00:36, krw@notreal.com wrote:
On Tue, 14 May 2019 21:45:36 +0100, Tom Gardner
spamjunk@blueyonder.co.uk> wrote:

On 14/05/19 17:00, John Larkin wrote:

snip

There is an optimum size of difficulty of problems that can be tackled
with today's available resources. So long as Moore's Law holds you can
prove that for some hard computational problems the fastest way to the
solution is to go surfing on the beach for a couple of years and then
start building your hardware using the latest fastest CPUs and memory.

That's like saying "I'll be later getting to work if I leave now,
rather than an hour from now because the traffic is worse this time of
the day.". Absurd.

No, it's like saying "should I jump on the bus now, or wait until my
wife returns from the shops so I can drive the car to work?"'

Bullshit

You really are dumber than a rock!

You're full of shit. But we all knew that.

And you are wilfully ignorant and stupid as well.

NO, you're simply illiterate.

Here is a concrete example taken from the time when the typical
improvement in clock speed and performance was roughly a 3 fold increase
for every 5 years elapsed. Once you start running the software you are
then committed to the technology that you begin to work with.

Absolutely irrelevant.

You clearly fail to understand. That is not my problem!

Not much at thinking, either.

Not the kind of thinking that krw can follow, perhaps. Krw's thinking (such as it is) starts with the proposition that whatever krw thinks is absolutely correct, and all that needs to be thought about the area under consideration.

Krw doesn't know much, and a lot of what he thinks he knows is over-simplified to the point of being misleading.

<snipped the stuff that krw didn't bother reading>

--
Bill Sloman, Sydney

 

[Rick C]

On Sunday, May 19, 2019 at 10:12:48 PM UTC-4, Bill Sloman wrote:

On Monday, May 20, 2019 at 11:44:08 AM UTC+10, k...@notreal.com wrote:
On Sun, 19 May 2019 16:55:35 +0100, Martin Brown
'''newspam'''@nezumi.demon.co.uk> wrote:

On 18/05/2019 04:45, krw@notreal.com wrote:
On Fri, 17 May 2019 08:41:50 +0100, Martin Brown
'''newspam'''@nezumi.demon.co.uk> wrote:

On 17/05/2019 02:31, krw@notreal.com wrote:
On Thu, 16 May 2019 14:03:12 +1000, Clifford Heath
no.spam@please.net> wrote:

On 16/5/19 1:08 pm, Bill Sloman wrote:
On Thursday, May 16, 2019 at 11:22:05 AM UTC+10, k...@notreal.com wrote:
On Wed, 15 May 2019 11:02:34 +0100, Martin Brown
'''newspam'''@nezumi.demon.co.uk> wrote:

On 15/05/2019 00:36, krw@notreal.com wrote:
On Tue, 14 May 2019 21:45:36 +0100, Tom Gardner
spamjunk@blueyonder.co.uk> wrote:

On 14/05/19 17:00, John Larkin wrote:

snip

There is an optimum size of difficulty of problems that can be tackled
with today's available resources. So long as Moore's Law holds you can
prove that for some hard computational problems the fastest way to the
solution is to go surfing on the beach for a couple of years and then
start building your hardware using the latest fastest CPUs and memory.

That's like saying "I'll be later getting to work if I leave now,
rather than an hour from now because the traffic is worse this time of
the day.". Absurd.

No, it's like saying "should I jump on the bus now, or wait until my
wife returns from the shops so I can drive the car to work?"'

Bullshit

You really are dumber than a rock!

You're full of shit. But we all knew that.

And you are wilfully ignorant and stupid as well.

NO, you're simply illiterate.

Here is a concrete example taken from the time when the typical
improvement in clock speed and performance was roughly a 3 fold increase
for every 5 years elapsed. Once you start running the software you are
then committed to the technology that you begin to work with.

Absolutely irrelevant.

You clearly fail to understand. That is not my problem!

Not much at thinking, either.

Not the kind of thinking that krw can follow, perhaps. Krw's thinking (such as it is) starts with the proposition that whatever krw thinks is absolutely correct, and all that needs to be thought about the area under consideration.

Krw doesn't know much, and a lot of what he thinks he knows is over-simplified to the point of being misleading.

snipped the stuff that krw didn't bother reading

This really is one bizarre collection of miscreants.

--

Rick C.

- Get 5,000 miles of free Supercharging
- Tesla referral code - https://ts.la/richard11209

 

[Bill Sloman]

On Friday, May 17, 2019 at 1:14:24 PM UTC+10, Bill Sloman wrote:

On Friday, May 17, 2019 at 11:31:38 AM UTC+10, k...@notreal.com wrote:
On Thu, 16 May 2019 14:03:12 +1000, Clifford Heath
no.spam@please.net> wrote:

On 16/5/19 1:08 pm, Bill Sloman wrote:
On Thursday, May 16, 2019 at 11:22:05 AM UTC+10, k...@notreal.com wrote:
On Wed, 15 May 2019 11:02:34 +0100, Martin Brown
'''newspam'''@nezumi.demon.co.uk> wrote:

On 15/05/2019 00:36, krw@notreal.com wrote:
On Tue, 14 May 2019 21:45:36 +0100, Tom Gardner
spamjunk@blueyonder.co.uk> wrote:

On 14/05/19 17:00, John Larkin wrote:

snip

There is an optimum size of difficulty of problems that can be tackled
with today's available resources. So long as Moore's Law holds you can
prove that for some hard computational problems the fastest way to the
solution is to go surfing on the beach for a couple of years and then
start building your hardware using the latest fastest CPUs and memory.

That's like saying "I'll be later getting to work if I leave now,
rather than an hour from now because the traffic is worse this time of
the day.". Absurd.

No, it's like saying "should I jump on the bus now, or wait until my
wife returns from the shops so I can drive the car to work?"'

Bullshit

Which means that krw can't follow the logic, which isn't surprising - the logic isn't demanding but krw's brain doesn't seem to work.

If you start work with a technology that slows you down, it's easy to
get too invested in it to be able to jump ship when a better way comes
along - so it's better not to start yet.

It doesn't "slow you down". There is nothing saying that you can't
change horses. Absurd.

Nobody says that you can't change horses, but you do have to pay a bundle for the extensive redesign required to accommodate new and better technology. It's not as expensive as designing from scratch, but it's still quite expensive enough to sink most projects.

snipped the stuff that krw didn't read about my experiences where I couldn't "change horses".

--
Bill Sloman, Sydney