Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

My prime counting function versus Legendre's

3 views
Skip to first unread message

James Harris

unread,
Jul 10, 2002, 9:51:04 PM7/10/02
to
It amazes me that so simple a lie as the assertion that my prime counting
function and Legendre's are the same is so effective.

Just consider that prime counting is such a well-worked and famous area in
mathematics, and there's such a small number of known difference equations
that a new prime counting function should be remarkable. In fact it should
be incredible given the amount of work done researching primes.

And then ponder the silence at my introduction of a new prime counting
difference function. I think it's just further proof that I've been right
all along about the "mathematical establishment". Look no further than what
I'm about to show you.

Luckily, it's easy to demonstrate how the two methods are different, while
the primary similarity is that they both give the same final answer (gee,
why would that surprise you?).

For instance, the prime count under 10 can be found using just the primes 2
and 3 with both methods.

(Reference: http://mathworld.wolfram.com/LegendresFormula.html)

With Legendre's (using "floor" instead of my usual "int"),

10 - floor(10/2) - floor(10/3) + floor(10/6) + pi(sqrt(10))- 1 gives

10 - 5 - 3 + 1 + 2-1 = 4.

While with my prime counting function,

5 - floor((10-4)/6) gives

5 - 1 = 4.

It gets worse, or should I say better?

Using 2, 3 and 5 lets us count with 26.

Legendres:

26 - floor(26/2) - floor(26/3) - floor(26/5) + floor(26/2(3)) + floor
(26/2(5)) + floor(26/3(5)) - floor(26/2(3)5) + pi(sqrt(26)) - 1, which is

26 - 13 - 8 - 5 + 4 + 2 +1 - 0 + 3 - 1 = 9

My prime counting function:

13 - floor((26-4)/6) - floor((26+4)/10) + floor((26 - 16)/30) + pi(4),
which is

13 - 3 - 3 + 0 + 2 = 9.

Need I go on?

How could *any* of you believe that the two are the same?

Could it be you *want* to believe, regardless of the truth?

At worse you should be able to admit that my prime counting function is
*shorter*. At best you might also admit that it looks *cleaner*.

So, why is this function being ignored?

Why are mathematicians teaching people Legendre's while ignoring a method
that's far simpler to express?

Because mathematicians like too many others on this planet are less about
truth and beauty in mathematics than they are about being political animals.

The political stakes here are HUGE. They are far bigger than most of you
probably realize, as mathematicians need you more than you realize. And
they need you to respect them as THE absolute experts when it comes to
mathematics. They want you to accept that NO valuable mathematics comes
from outside the establishment--their world.

They are the gatekeepers. And they have decreed that professional
mathematicians, or at least professionally trained mathematicians must use
their gateways or suffer--no matter how beautiful or groundbreaking the
mathematics is. I'm sure your knee-jerk reaction is to *believe* I'm wrong.
After all, I'm sure that makes you feel comfortable.

Need I remind you of the things going on in this world that are true
*despite* your comfort level.

The US economy sags, while people finally admit that corporate officers lie
and cheat to make themselves rich at the expense of stockholders (gee, why
is that a shocker?).

While most people discounted AIDS, it's still out there with a vengeance,
and the worst is probably yet to come. But I'm sure you have your own
truth, that you prefer, eh?

Catholics are still dealing with the aftershocks of all the revelations of
priests who molested children. But I'm sure you all would have certainly
believe anyone who told you what was going on early on, eh?

It was announced today that humanity is using up the world's resources at a
rate likely to fundamentally deplete them by 2050. But hey, that's a while
yet, now isn't it. I'm in America, so I help consume more than most of you
anyway, so gee, why should I care about that, eh?

Recent research shows that hormone replacement therapy for women does more
harm than good, yet doctors will probably still prescribe because as one
doctor said in an article I read today (I think it was The New York Times)
(slight paraphrase), "truth is a matter of opinion."

So when truth is a matter of opinion, and mathematicians would just as soon
not accept a truth from someone not acceptable to them, by those rules, no
matter how beautiful, useful, or incredible a new prime counting function
might be, they WILL NOT ADMIT IT!!!

It gets still worse.

As I've mentioned in other posts, my prime counting function actually points
the way to directly calculating the continuous pi function, which is also
why the *political* stakes are so huge.

You may have heard of the Riemann Hypothesis. It's a guess made by one of
the greatest mathematicians ever (the guy who came up with the math used by
Einstein for his General Theory of Relativity) about an equation that's a
step removed from what I'm talking about.

That is, if Riemann had my prime counting function, he'd probably never have
bothered with the equation he made his hypothesis about.

That's why it's so important for mathematicians to keep me under wraps, and
it is about keeping me under wraps.

After all, far worse than just being some amateur who plays at mathematics
as a hobby, I've been very vocal in criticising mathematicians. They can
only expect worse if I get a worldwide stage with any kind of validation.

But, that's the kind of thing you might expect from politicians NOT from
scientists.

What makes it truly sad is that I'm reduced to making these kind of pleas
with the hopes that there are some of you out there who care less about
mathematician's need to feel like they control all mathematics than you care
about knowing the truth.

After all, think the next time when you ponder Riemann's Hypothesis, how
pathetic it will be for you to spend the time wondering about the
complexity, trying to get a grasp, if there's a far simpler answer just
beyond your reach, only because a minority of men and women who *call*
themselves mathematicians wish it so.

You see, you lose, and what do they really win?

A world where humanity is dumber than it thinks it is, just because a group
of people wish to be considered smarter than they really are.


James Harris


JSH Info

unread,
Jul 10, 2002, 10:03:00 PM7/10/02
to

For information about James Harris see:
http://www.geocities.com/williamrexmarshall/cranks/jsh1.html

SURGEON GENERAL'S WARNING:
"Try not to respond to James.
Together we can do it."

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Last week there were comments from various posters
regarding problems mainly with your Area One proof. You
have failed to respond to these, preferring evidently to
go out and do some drinking, and then returning to say
you were worried about being schizophrenic, etc.. These
critical (and mathematical) postings have now slipped
below the sci.math horizon, but you still have not
responded and your web site has not changed. Instead you
are back to your usual unseemly bragging, whining, and
dissing of mathematicians. When are you going to get
back to concentrating on the math?
-- The above is part of a post by Nora Baron to sci.math.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

The primary purpose of these posts is to give
newbies a pointer to some background information on James.
A message similar to this one is automatically
posted to every thread in sci.math started by him.
It appears that he starts a new thread whenever he considers
himself to be losing the argument in a previous one.
This is automatic post number 252.
New threads started by James in the last 7 days: 8 (1.1 threads/day)
New threads started by James in the last 30 days: 33 (1.1 threads/day)
New threads started by James in the last 90 days: 99 (1.1 threads/day)
James: It would be nice if you included JSH in the subject.

Some news readers allow the user to filter out
postings with a particular phrase in the subject.

--
JSH Info, version 1.0.0.234

Paul Sperry

unread,
Jul 10, 2002, 11:58:10 PM7/10/02
to
In article <#rKoa1HKCHA.1352@cpimsnntpa03>,
"James Harris" <jst...@msn.com> wrote:

>

L. Carroll:
"He only does it to annoy
Because he knows it teases"

> It amazes me that so simple a lie as the assertion that my prime counting
> function and Legendre's are the same is so effective.

Can an assertion be a lie? Anyway, as I'm sure you know, "your" prime
counting function _is_ the same as Legendre's. In fact it is the same as
_my_ prime counting function.

[...]

> Luckily, it's easy to demonstrate how the two methods are different,

See there! I knew you knew it was the methods, i.e. algorithms, which
are, ostensibly, different not the functions.

[...]

> So, why is this function being ignored?

Because most are just not interested. I wonder if you have any idea how
_little_ interest the mathematics community has in the prime counting
function.... For good reason.



> Why are mathematicians teaching people Legendre's while ignoring a method
> that's far simpler to express?

They're not. You're in good company; they're ignoring Legendre's method
too. In the vast majority of classes the prime counting function is
never mentioned.

> Because mathematicians like too many others on this planet are less about
> truth and beauty in mathematics than they are about being political animals.

Mathematicians?! Political animals?!! Bwahahahahahaha!



> The political stakes here are HUGE.

Good one!

[...]

< Chuckle, chuckle > Sorry, I'm still contemplating mathematicians as
political animals.

[...]

> As I've mentioned in other posts, my prime counting function actually points
> the way to directly calculating the continuous pi function, which is also
> why the *political* stakes are so huge.

Since, as far as I know, no one has asked this, I will. In what sense is
the prime counting function continuous?

[...]

> That is, if Riemann had my prime counting function, he'd probably never have
> bothered with the equation he made his hypothesis about.

And if frogs had wings they probably wouldn't have to bump their butts
in the mud.

[...]

> After all, think the next time when you ponder Riemann's Hypothesis,

It's astonishing how little time I spend contemplating the Riemann
Hypothesis. The next time I do, I'll be sure to think of you.

[...]

> James Harris
--
Paul Sperry
Columbia, SC (USA)

Franz Fritsche

unread,
Jul 11, 2002, 12:54:29 AM7/11/02
to
On Thu, 11 Jul 2002 03:58:10 GMT, Paul Sperry <plsp...@sc.rr.com>
wrote:

>>
>> (...) mathematicians like too many others on this planet are less about


>> truth and beauty in mathematics than they are about being political animals.
>>
> Mathematicians?! Political animals?!! Bwahahahahahaha!
>

> [...]
>
> < Chuckle, chuckle > Sorry, I'm still contemplating mathematicians as
> political animals.
>

Well, the joke is not as big a joke as you might think... :-)

Since Aristotle said each person is a "Zoon Politicon". This
is often translated "Political Animal".

"According to Aristotle human beings are essentially social. In this
context we should interpret Aristotle's claim that a human being is a
zoon politicon. In direct translation from Greek zoon politicon means
"a political animal [living being --ff]". What Aristotle meant by this
expression was that we can be distinguished from all other animals by
the fact that we form a political society -- a society in which we can
establish a certain rational order."

But Aristotle also defined the human being as "Zoon Logistikon"
("animal rationale").

So maybe what JSH (our most political animal here in sci.math)
probably wants to say, is:

"mathematicians like too many others on this planet are less about

truth and beauty in mathematics (animal rationale) than they are about
being political animals (animal sociale)."

Now for Aristotle this was NO contradiction!


More about Aristotle's notation of human being a "Zoon Politicon":

"Aristotle observed that our virtues, character as well as our dreams
and ambitions are shaped by the society. More importantly, most of our
aspirations and projects do not make sense outside of the context of
human community. For instance, you cannot become a college graduate,
win the Olympics or even do your shopping if there is no society that
makes these functions possible. Furthermore, you cannot speak about
virtues, like courage or beneficence, if there is no community that
you want to help with your actions.

Aristotle's conception of human nature has its applications in his
understanding of the good life. According to Aristotle, one cannot
have a good life if one does not have good friends. Moreover, in order
to have a good life a person needs to belong to various good
communities starting from one's family -- all the way to the broadest
community, which for Aristotle is one's country. This gives every
member of the community an incentive to improve the state as a
necessary means to improving one's own life."

F.

LarryLard

unread,
Jul 11, 2002, 5:06:04 AM7/11/02
to
"James Harris" <jst...@msn.com> wrote in message news:<#rKoa1HKCHA.1352@cpimsnntpa03>...
[snippage]

> It amazes me that so simple a lie as the assertion that my prime counting
> function and Legendre's are the same is so effective.

If your formula is correct, then it describes the same function as
Legndre's does. You are still using the wrong word, and I think it's
because you think you actually have a different function. Tell me, do
you think these two functions are different? :

f: N -> N, f(n) = 2n

g: N -> N, g(n) = n+n

> Just consider that prime counting is such a well-worked and famous area in
> mathematics, and there's such a small number of known difference equations
> that a new prime counting function should be remarkable. In fact it should
> be incredible given the amount of work done researching primes.

By the way, have you proved your formula's correctness?

> With Legendre's (using "floor" instead of my usual "int"),

You'll find floor more common than int in mathematics, as opposed to
computer programming.

> Why are mathematicians teaching people Legendre's while ignoring a method
> that's far simpler to express?

Because one has been proved correct(*) while the other hasn't?

(*) I'm assuming here that Legendre's formula has been proved correct


> The political stakes here are HUGE.

*This* old chestnut.

> They are far bigger than most of you
> probably realize, as mathematicians need you more than you realize. And
> they need you to respect them as THE absolute experts when it comes to
> mathematics. They want you to accept that NO valuable mathematics comes
> from outside the establishment--their world.
>
> They are the gatekeepers.

Yes, and you are the keymaster. Or something. Beware Gozer the
Destructor!

> Need I remind you of the things going on in this world that are true
> *despite* your comfort level.

[...]


> So when truth is a matter of opinion, and mathematicians would just as soon
> not accept a truth from someone not acceptable to them, by those rules, no
> matter how beautiful, useful, or incredible a new prime counting function
> might be, they WILL NOT ADMIT IT!!!

So, let me get this straight. Because there exist facts that people
are unwilling to accept, this means that anything people are unwilling
to accept is a fact? So similarly, because there exist horses that are
brown, does that make all brown things horses? This logical fallacy is
as old as Socrates!

People are also unwilling to accept that I have a divine right to
dominion over all the nations of the world. Folks, this is HUGE!

> It gets still worse.

Don't we know it.

> As I've mentioned in other posts, my prime counting function actually points
> the way to directly calculating the continuous pi function

The what?

> , which is also
> why the *political* stakes are so huge.
>
> You may have heard of the Riemann Hypothesis.

Some of us can even state it. Can you?

> It's a guess made by one of
> the greatest mathematicians ever (the guy who came up with the math used by
> Einstein for his General Theory of Relativity) about an equation that's a
> step removed from what I'm talking about.

Which equation would that be?

> That is, if Riemann had my prime counting function, he'd probably never have
> bothered with the equation he made his hypothesis about.

We can all play 'what if'.

> That's why it's so important for mathematicians to keep me under wraps, and
> it is about keeping me under wraps.

If only we could!

> After all, far worse than just being some amateur who plays at mathematics
> as a hobby, I've been very vocal in criticising mathematicians. They can
> only expect worse if I get a worldwide stage with any kind of validation.

Now *there's* a big if.

> What makes it truly sad is that I'm reduced to making these kind of pleas
> with the hopes that there are some of you out there who care less about
> mathematician's need to feel like they control all mathematics than you care
> about knowing the truth.

So - and let me get this straight - you're still looking for even ONE
person who supports you?


--
Larry Lard
Replies to group please.

fuffy

unread,
Jul 11, 2002, 5:23:23 AM7/11/02
to
"James Harris" <jst...@msn.com> wrote in message news:<#rKoa1HKCHA.1352@cpimsnntpa03>...
> So, why is this function being ignored?
>
> Why are mathematicians teaching people Legendre's while ignoring a method
> that's far simpler to express?
>

My... BUTT! My... BUTT! I... can't.... find... my... BUTT!!!!!!

WilmaScranton

unread,
Jul 11, 2002, 8:33:04 AM7/11/02
to
From: "James Harris" jst...@msn.com

However, the time taken to compute pi(n) by your method still appears to be
linear in n and thus probably unsuitable for computing pi(n) for large n. I got
the following timings:

pi( 100000000)= 5761455 Time: 5168
pi( 200000000)=11078937 Time: 10276
pi( 400000000)=21336326 Time: 20470
pi( 800000000)=41146179 Time: 40841
pi(1600000000)=79451833 Time: 82201

I had taken out your diagnostic displays.
The times are in milliseconds.

I also tried the algorithm with the lines like
if ((p2.pi(2*j+4) - p2.pi(2*j+2))==1){
replaced by a simple test for the primality of 2*j+3.
The program runs somewhat faster then, but not very much so.

By the way, Mathematica running on the same computer evalutes PrimePi[16*10^8]
as 79451833 (for which you use 82201 milliseconds) in about 40 milliseconds.

Also, as has been mentioned already, your algorithm gives such wrong results as
pi(3)=1 and pi(5)=2.

Wilma.

David C. Ullrich

unread,
Jul 11, 2002, 10:25:19 AM7/11/02
to
On Wed, 10 Jul 2002 21:51:04 -0400, "James Harris" <jst...@msn.com>
wrote:

>It amazes me that so simple a lie as the assertion that my prime counting
>function and Legendre's are the same is so effective.

That's a lie, all right. They're clearly not the same: Legendre's
formula gives the right answer for pi(3).

>Just consider that prime counting is such a well-worked and famous area in
>mathematics, and there's such a small number of known difference equations
>that a new prime counting function should be remarkable. In fact it should
>be incredible given the amount of work done researching primes.
>
>And then ponder the silence at my introduction of a new prime counting
>difference function.

What silence? When it became clear that it was more or less right,
sort of, people said they were impressed.

And when you started claiming it was the fastest possible prime-
counting algorithm there was no silence - there were loud derisive
jeers, as appropriate for statements that are so _easily_ shown to
be false. Similarly with your claims that it shows that you're
in the same league with Gauss and Dedekind, your claims that
the prime counting function is going to have huge consequences
in everyday life, your weird statements about using a bouncing
ball to count primes, your claims that the "function" can be
used to settle the Riemann Hypothesis, etc.

> I think it's just further proof that I've been right
>all along about the "mathematical establishment".

Well, I don't know about "further" proof. Your hundreds of
previous proofs the you were right about the mathematical
establishment were based on said estanlishments rejection
of things that _you_ eventually admitted were wrong, so
they turned out to be proofs that the mathematical establishment
is doing pretty well.

But it's good that we _finally_ have a valid proof that you're
right about the mathematical establishment.

>Look no further than what
>I'm about to show you.
>
>Luckily, it's easy to demonstrate how the two methods are different, while
>the primary similarity is that they both give the same final answer (gee,
>why would that surprise you?).

It would surprise us if they gave the same answer because so many of
the things you've done have been nonsense. But in fact it doesn't
give the same answer, no surprise.

[...]


>
>Need I go on?
>
>How could *any* of you believe that the two are the same?
>
>Could it be you *want* to believe, regardless of the truth?
>
>At worse you should be able to admit that my prime counting function is
>*shorter*. At best you might also admit that it looks *cleaner*.

I don't think that anyone would agree that there's anything "clean"
about an implementation of pi(n) that checks whether x is prime
using recursive calls to pi(x)-pi(x-1)...

>So, why is this function being ignored?

Maybe because much faster methods are known? No, that's too simple.

>Why are mathematicians teaching people Legendre's while ignoring a method
>that's far simpler to express?

I've never seen anyone teach Legendre's algorithm.

>Because mathematicians like too many others on this planet are less about
>truth and beauty in mathematics than they are about being political animals.
>
>The political stakes here are HUGE. They are far bigger than most of you
>probably realize, as mathematicians need you more than you realize. And
>they need you to respect them as THE absolute experts when it comes to
>mathematics. They want you to accept that NO valuable mathematics comes
>from outside the establishment--their world.
>
>They are the gatekeepers. And they have decreed that professional
>mathematicians, or at least professionally trained mathematicians must use
>their gateways or suffer--no matter how beautiful or groundbreaking the
>mathematics is. I'm sure your knee-jerk reaction is to *believe* I'm wrong.
>After all, I'm sure that makes you feel comfortable.
>
>Need I remind you of the things going on in this world that are true
>*despite* your comfort level.
>
>The US economy sags, while people finally admit that corporate officers lie
>and cheat to make themselves rich at the expense of stockholders (gee, why
>is that a shocker?).
>
>While most people discounted AIDS, it's still out there with a vengeance,
>and the worst is probably yet to come. But I'm sure you have your own
>truth, that you prefer, eh?
>
>Catholics are still dealing with the aftershocks of all the revelations of
>priests who molested children. But I'm sure you all would have certainly
>believe anyone who told you what was going on early on, eh?

If only we'd believe that an algorithm that takes _much_ longer than
Mathematica is the fastest possible algorithm, that would solve the
problems with AIDS and priests diddling little boys. Oh well, it's
still not going to happen.

>It was announced today that humanity is using up the world's resources at a
>rate likely to fundamentally deplete them by 2050. But hey, that's a while
>yet, now isn't it. I'm in America, so I help consume more than most of you
>anyway, so gee, why should I care about that, eh?
>
>Recent research shows that hormone replacement therapy for women does more
>harm than good, yet doctors will probably still prescribe because as one
>doctor said in an article I read today (I think it was The New York Times)
>(slight paraphrase), "truth is a matter of opinion."
>
>So when truth is a matter of opinion, and mathematicians would just as soon
>not accept a truth from someone not acceptable to them, by those rules, no
>matter how beautiful, useful, or incredible a new prime counting function
>might be, they WILL NOT ADMIT IT!!!
>
>It gets still worse.

Oh my.

>As I've mentioned in other posts, my prime counting function actually points
>the way to directly calculating the continuous pi function, which is also
>why the *political* stakes are so huge.
>
>You may have heard of the Riemann Hypothesis. It's a guess made by one of
>the greatest mathematicians ever (the guy who came up with the math used by
>Einstein for his General Theory of Relativity) about an equation that's a
>step removed from what I'm talking about.
>
>That is, if Riemann had my prime counting function, he'd probably never have
>bothered with the equation he made his hypothesis about.

Oops. When you imply that your Function is the key to curing AIDS I
can buy that. But here you're back to babbling nonsense.

>That's why it's so important for mathematicians to keep me under wraps, and
>it is about keeping me under wraps.

Exactly how is it that anyone's been keeping you under wraps? You've
posted thousands of posts on the internet - nobody's prevented you
from doing that. Editors have refused to publish papers of yours -
in both cases you finally admitted that the results that they
refused to publish were wrong.

>After all, far worse than just being some amateur who plays at mathematics
>as a hobby, I've been very vocal in criticising mathematicians. They can
>only expect worse if I get a worldwide stage with any kind of validation.
>
>But, that's the kind of thing you might expect from politicians NOT from
>scientists.
>
>What makes it truly sad is that I'm reduced to making these kind of pleas
>with the hopes that there are some of you out there who care less about
>mathematician's need to feel like they control all mathematics than you care
>about knowing the truth.
>
>After all, think the next time when you ponder Riemann's Hypothesis, how
>pathetic it will be for you to spend the time wondering about the
>complexity, trying to get a grasp, if there's a far simpler answer just
>beyond your reach, only because a minority of men and women who *call*
>themselves mathematicians wish it so.

It would be good if you said exactly what the connection between your
algorithm and RH _was_, instead of all these hints that there's a
link.

>You see, you lose, and what do they really win?
>
>A world where humanity is dumber than it thinks it is, just because a group
>of people wish to be considered smarter than they really are.

And if we'd only give in and acknowledge your transcendent genius,
that would be a good thing, Truth-triumphing-over-Ignorance-wise.
Right.

>James Harris
>


David C. Ullrich

David C. Ullrich

unread,
Jul 11, 2002, 10:28:06 AM7/11/02
to
On 11 Jul 2002 02:23:23 -0700, rampant_f...@yahoo.co.uk (fuffy)
wrote:

The typical reason for this is that you're sitting on it. HTH.

David C. Ullrich

James Harris

unread,
Jul 11, 2002, 11:11:19 AM7/11/02
to
Paul Sperry <plsp...@sc.rr.com> wrote in message news:<plsperry-0FFBD1...@news-server.sc.rr.com>...

> In article <#rKoa1HKCHA.1352@cpimsnntpa03>,
> "James Harris" <jst...@msn.com> wrote:
>
> >
>
> L. Carroll:
> "He only does it to annoy
> Because he knows it teases"
>
> > It amazes me that so simple a lie as the assertion that my prime counting
> > function and Legendre's are the same is so effective.
>
> Can an assertion be a lie? Anyway, as I'm sure you know, "your" prime
> counting function _is_ the same as Legendre's. In fact it is the same as
> _my_ prime counting function.
>
> [...]
>
> > Luckily, it's easy to demonstrate how the two methods are different,
>
> See there! I knew you knew it was the methods, i.e. algorithms, which
> are, ostensibly, different not the functions.
>
> [...]
>
> > So, why is this function being ignored?
>
> Because most are just not interested. I wonder if you have any idea how
> _little_ interest the mathematics community has in the prime counting
> function.... For good reason.

LOL. The Riemann Hypothesis is the single phrase I produce to raise
the strong possibility that you apparently think that mathematicians
can go from "spin" to outright lying.

The Riemann Hypothesis is just a very informed guess about the nature
of the distribution of prime numbers i.e. about the prime counting
function.

Ok folks, what does it tell you when people go to the trouble of
making posts like this one where they can get slammed so easily?

I don't know. Guessing about such things is boring.

Then again, maybe I should be fairer. Possibly this fellow is talking
about *difference* prime counting functions, which the mathematicians
seem to have *thought* they knew all they needed to know about.

Remember folks, mathematics is an infinite subject. Just because the
people who today call themselves mathematicians think they have all
the good stuff in a particular area figured out, doesn't mean they
actually DO have it all figured it out. It's actually somewhat
moronic--these people will TELL you that nothing important can be in
an area because none of them found it!!!

How do I know that?

Because I've been told that repeatedly.

So mathematicians couldn't find any difference equations for counting
primes that lead to a continuous function. (So why should I care?
Why should my work be ignored because they couldn't find it first? Is
that fair? Of course it's not fair, or sensible.)

That's why Riemann went to the Euler power formula in the first place
because they didn't know how to get to the continuous prime function
any other way.

Well I just showed the other way, the best way.

So, is this guy truly clueless? Or is he lying to you?

I don't know. Guessing about such things is boring.


James Harris

Phil Carmody

unread,
Jul 11, 2002, 12:42:39 PM7/11/02
to
WilmaScranton wrote:
> From: "James Harris" jst...@msn.com

> >With Legendre's (using "floor" instead of my usual "int"),
> >
> > 10 - floor(10/2) - floor(10/3) + floor(10/6) + pi(sqrt(10))- 1 gives
> >
> > 10 - 5 - 3 + 1 + 2-1 = 4.
> >
> >While with my prime counting function,
> >
> > 5 - floor((10-4)/6) gives
> >
> > 5 - 1 = 4.
> >
> >It gets worse, or should I say better?
> >
> >Using 2, 3 and 5 lets us count with 26.
> >
> >Legendres:
> >
> > 26 - floor(26/2) - floor(26/3) - floor(26/5) + floor(26/2(3)) + floor
> >(26/2(5)) + floor(26/3(5)) - floor(26/2(3)5) + pi(sqrt(26)) - 1, which is
> >
> > 26 - 13 - 8 - 5 + 4 + 2 +1 - 0 + 3 - 1 = 9
> >
> >My prime counting function:
> >
> > 13 - floor((26-4)/6) - floor((26+4)/10) + floor((26 - 16)/30) + pi(4),
> >which is
> >
> > 13 - 3 - 3 + 0 + 2 = 9.
> >
> >Need I go on?

It becomes likely that James's 'tweak' is a method of pairing terms in
the Legendre sum.

> However, the time taken to compute pi(n) by your method still appears to be
> linear in n and thus probably unsuitable for computing pi(n) for large n. I got
> the following timings:
>
> pi( 100000000)= 5761455 Time: 5168
> pi( 200000000)=11078937 Time: 10276
> pi( 400000000)=21336326 Time: 20470
> pi( 800000000)=41146179 Time: 40841
> pi(1600000000)=79451833 Time: 82201
>
> I had taken out your diagnostic displays.
> The times are in milliseconds.

Thank you Wilma for that quick test. It would be interesting to see how
well a similarly coded bread-and-butter Legendre sum would run. I would
be prepared to accept that James may have found a way to double its
speed by this term-pairing. However, in the end that's just a constant
factor speedup.



> I also tried the algorithm with the lines like
> if ((p2.pi(2*j+4) - p2.pi(2*j+2))==1){
> replaced by a simple test for the primality of 2*j+3.
> The program runs somewhat faster then, but not very much so.

Legendre's permitted to use O(x^0.5) space, so a cached sieved range is
well within the rules. I'm surprised that the above change makes so
little difference.

> By the way, Mathematica running on the same computer evalutes PrimePi[16*10^8]
> as 79451833 (for which you use 82201 milliseconds) in about 40 milliseconds.

Mathematica "cheats" though. It has precomputed values for pi, and maybe
only needs to sieve end-zones.



> Also, as has been mentioned already, your algorithm gives such wrong results as
> pi(3)=1 and pi(5)=2.

I've heard this repeated so many times. Why doesn't he just add
something like
if(n<6)return t[n];


Phil
(£16)

Andy Spragg

unread,
Jul 11, 2002, 4:59:53 PM7/11/02
to
David C. Ullrich <ull...@math.okstate.edu> pushed briefly to the
front of the queue on Thu, 11 Jul 2002 09:28:06 -0500, and nailed this
to the shed door:

^ On 11 Jul 2002 02:23:23 -0700, rampant_f...@yahoo.co.uk (fuffy)
^ wrote:
^
^ >"James Harris" <jst...@msn.com> wrote in message news:<#rKoa1HKCHA.1352@cpimsnntpa03>...
^ >> So, why is this function being ignored?
^ >>
^ >> Why are mathematicians teaching people Legendre's while ignoring a method
^ >> that's far simpler to express?
^ >>
^ >
^ >My... BUTT! My... BUTT! I... can't.... find... my... BUTT!!!!!!
^
^ The typical reason for this is that you're sitting on it. HTH.

Unless you're JSH, when the typical reason is that you have your head
up it.

Andy

--
sparge at globalnet point co point uk

Life begins at kilofortnight

Lynn Killingbeck

unread,
Jul 11, 2002, 6:03:58 PM7/11/02
to
James Harris wrote:
>
> (snip)

>
> Ok folks, what does it tell you when people go to the trouble of
> making posts like this one where they can get slammed so easily?
>
> I don't know. Guessing about such things is boring.
>
> (snip)
>
> James Harris

Why, that's easy! They tell me the post was by JSH! No guessing needed!

Lynn Killingbeck

P.S. Sorry, I mis-clicked into this JSH thread, and couldn't resist.

David Kastrup

unread,
Jul 11, 2002, 7:10:24 PM7/11/02
to
spa...@globalnet.co.uk (Andy Spragg) writes:

Why "unless"? Some of his posts would make it quite likely that he
has been sitting on his head while there. Unhealthy compression.

--
David Kastrup, Kriemhildstr. 15, 44793 Bochum
Email: David....@t-online.de

Dik T. Winter

unread,
Jul 11, 2002, 9:31:22 PM7/11/02
to
> WilmaScranton wrote:

In article <3D2DB57F...@yahoo.co.uk> Phil Carmody <thefatphil_d...@yahoo.co.uk> writes:
> > From: "James Harris" jst...@msn.com
> > >With Legendre's (using "floor" instead of my usual "int"),
> > > 10 - floor(10/2) - floor(10/3) + floor(10/6) + pi(sqrt(10))- 1 gives
> > > 10 - 5 - 3 + 1 + 2-1 = 4.
> > >While with my prime counting function,
> > > 5 - floor((10-4)/6) gives
> > > 5 - 1 = 4.
> > >It gets worse, or should I say better?
...

> It becomes likely that James's 'tweak' is a method of pairing terms in
> the Legendre sum.

No. He just considers odd numbers only.
--
dik t. winter, cwi, kruislaan 413, 1098 sj amsterdam, nederland, +31205924131
home: bovenover 215, 1025 jn amsterdam, nederland; http://www.cwi.nl/~dik/

Paul Sperry

unread,
Jul 12, 2002, 1:25:40 AM7/12/02
to
In article <3c65f87.02071...@posting.google.com>,
jst...@msn.com (James Harris) wrote:

> Paul Sperry <plsp...@sc.rr.com> wrote in message
> news:<plsperry-0FFBD1...@news-server.sc.rr.com>...
> > In article <#rKoa1HKCHA.1352@cpimsnntpa03>,
> > "James Harris" <jst...@msn.com> wrote:

> > L. Carroll:
> > "He only does it to annoy
> > Because he knows it teases"

[...]

> > > So, why is this function being ignored?
> >
> > Because most are just not interested. I wonder if you have any idea how
> > _little_ interest the mathematics community has in the prime counting
> > function.... For good reason.
>
> LOL. The Riemann Hypothesis is the single phrase I produce to raise
> the strong possibility that you apparently think that mathematicians
> can go from "spin" to outright lying.

I'm sorry. I didn't follow that at all. My point is that I can be
disinterested in the prime counting function in the same sense that I am
disinterested in, say, origami. There are many no doubt interesting
things in which _I_ am not interested. There has never been an instance
where I needed or wanted to know how many primes were smaller than some
natural number.

Scenario; [Prof. A and Prof. B in the coffee room]
Prof. A: I just saw that some guy has a nifty algorithm for the prime
counting function.
Prof. B: Remind me what the prime counting function is.
Prof. A: [Tells him]
Prof. B: Good for him. How's the team going to do this week?

> The Riemann Hypothesis is just a very informed guess about the nature
> of the distribution of prime numbers i.e. about the prime counting
> function.

Well, RH is perhaps asymptotic to prime counting.

[...]

> Possibly this fellow is talking
> about *difference* prime counting functions, which the mathematicians
> seem to have *thought* they knew all they needed to know about.

I promise you this fellow was _not_ talking about "difference prime
counting functions". I have no idea what those are but I'm pretty sure I
wasn't talking, or even writing, about them.

> Remember folks, mathematics is an infinite subject. Just because the
> people who today call themselves mathematicians think they have all
> the good stuff in a particular area figured out, doesn't mean they
> actually DO have it all figured it out. It's actually somewhat
> moronic--these people will TELL you that nothing important can be in
> an area because none of them found it!!!

Can you give even one example where someone has said that? ... I thought
not.



> How do I know that?

Good question.

> Because I've been told that repeatedly.

Seems unlikely. By whom? Reference?

> So mathematicians couldn't find any difference equations for counting
> primes that lead to a continuous function. (So why should I care?
> Why should my work be ignored because they couldn't find it first? Is
> that fair? Of course it's not fair, or sensible.)
>
> That's why Riemann went to the Euler power formula in the first place
> because they didn't know how to get to the continuous prime function
> any other way.

I'll repeat my question. What is the _continuous_ prime counting
function?

> Well I just showed the other way, the best way.
>
> So, is this guy truly clueless? Or is he lying to you?

Are those our only choices? Well, let's see; I know there's a clue
around here somewhere. Ah, there it is. I'm curious about exactly what
it is that you think I am lying about.

More specifically what did I write that you feel is untrue and what is
your evidence that what I wrote was false?

Why do I not expect an answer to this?

Tris

unread,
Jul 12, 2002, 6:23:05 AM7/12/02
to

"David Kastrup" <David....@t-online.de> wrote in message
news:x5k7o1x...@tupik.goethe.zz...
> spa...@globalnet.co.uk (Andy Spragg) writes:
>
<snip>

> > ^ >My... BUTT! My... BUTT! I... can't.... find... my... BUTT!!!!!!
> > ^
> > ^ The typical reason for this is that you're sitting on it. HTH.
> >
> > Unless you're JSH, when the typical reason is that you have your head
> > up it.
>
> Why "unless"? Some of his posts would make it quite likely that he
> has been sitting on his head while there. Unhealthy compression.

That reminds me:
http://www.mazepath.com/uncleal/sunshine.jpg


Chris Thompson

unread,
Jul 12, 2002, 12:03:11 PM7/12/02
to
In article <plsperry-E29E93...@news-server.sc.rr.com>,
Paul Sperry <plsp...@sc.rr.com> wrote:
[...]

>
>I'm sorry. I didn't follow that at all. My point is that I can be
>disinterested in the prime counting function in the same sense that I am
>disinterested in, say, origami. There are many no doubt interesting
>things in which _I_ am not interested. There has never been an instance
>where I needed or wanted to know how many primes were smaller than some
>natural number.
>
>Scenario; [Prof. A and Prof. B in the coffee room]
> Prof. A: I just saw that some guy has a nifty algorithm for the prime
> counting function.
> Prof. B: Remind me what the prime counting function is.
> Prof. A: [Tells him]
> Prof. B: Good for him. How's the team going to do this week?

Gnaaargh! That's not "disinterested", it's "uninterested". This is
"disinterested":

Scenario; [Prof. A and Prof. B in the coffee room]
Prof. A: I just saw that some guy has a nifty algorithm for the prime
counting function.

Prof. B: Is it faster than Prof. C's, then?
Prof. A: I'm not too sure about that.
Prof. B: Would you like me to get Lowly Research Student D to code
it up and run it through our benchmarks?
Prof. A: Good idea. Better not tell Prof. C we are doing that just
yet, though...

Of course, in the current instance, this is followed some time later
by:

LoRS. D: I'm sorry, Professor, but I just haven't been able
to make any sense out of the description of this guy's algorithm.
Prof. B: Well, I'm not surprised. Prof. A is a bit gullible about
these things... don't tell anyone I said that! How's the team

going to do this week?

Chris Thompson
Email: cet1 [at] cam.ac.uk

Paul Sperry

unread,
Jul 12, 2002, 12:26:32 PM7/12/02
to
In article <agmujv$hfv$1...@pegasus.csx.cam.ac.uk>,
cet1-...@cam.ac.uk.invalid (Chris Thompson) wrote:

> In article <plsperry-E29E93...@news-server.sc.rr.com>,
> Paul Sperry <plsp...@sc.rr.com> wrote:
>
[...]
> >

[...]


> >I can be
> >disinterested in the prime counting function in the same sense that I am
> >disinterested in, say, origami.

[...]


> Gnaaargh! That's not "disinterested", it's "uninterested".

[...]

So, if I _did_ work up some enthusiasm for origami, I would soon
discover that I was misinterested so I would become deinterested leaving
me noninterrested. Have I got that right?

Chip Eastham

unread,
Jul 12, 2002, 1:53:02 PM7/12/02
to
rampant_f...@yahoo.co.uk (fuffy) wrote in message news:<db99d8aa.02071...@posting.google.com>...

Actually, I was thinking:

My... 'BOT! My... 'BOT! I... can't.... read... my... 'BOT!!!!!

Did Google just miss it on this thread, or is the BOT not visible on
other news servers for this thread as well?

-- chip

WilmaScranton

unread,
Jul 12, 2002, 2:51:02 PM7/12/02
to
From: eas...@bellsouth.net (Chip Eastham)

>> My... BUTT! My... BUTT! I... can't.... find... my... BUTT!!!!!!
>
>Actually, I was thinking:
>
>My... 'BOT! My... 'BOT! I... can't.... read... my... 'BOT!!!!!
>
>Did Google just miss it on this thread, or is the BOT not visible on
>other news servers for this thread as well?

It made it to AOL and to MathForum...
http://mathforum.org/epigone/sci.math/smayjinphil

It showed James at his most consistent!
1.1 posts per day in each time period...

========================


This is automatic post number 252.

New threads started by James in the last 7 days: 8 (1.1 threads/day)
New threads started by James in the last 30 days: 33 (1.1 threads/day)
New threads started by James in the last 90 days: 99 (1.1 threads/day)

========================

Wilma.

Dann Corbit

unread,
Jul 12, 2002, 2:32:35 PM7/12/02
to
"Paul Sperry" <plsp...@sc.rr.com> wrote in message
news:plsperry-49A161...@news-server.sc.rr.com...


Look everybody, pi(n) calculations are interesting to mathematicians. If
not, they ought to be. If Dik Winter or David Kastrup or someone like that
had shared the algorithm with us, we would all be keenly enjoying it (of
course, it would have been a lot more polished and the complexity would have
been correctly understood...).

At any rate, these cries of:
"Calculating pi(x) is dull, dull, dull!"
are only because it is James.

IMO-YMMV.
--
C-FAQ: http://www.eskimo.com/~scs/C-faq/top.html
"The C-FAQ Book" ISBN 0-201-84519-9
C.A.P. FAQ: ftp://cap.connx.com/pub/Chess%20Analysis%20Project%20FAQ.htm


James Harris

unread,
Jul 12, 2002, 7:48:28 PM7/12/02
to

"Dann Corbit" <dco...@connx.com> wrote in message
news:agn78...@enews2.newsguy.com...

Yup, they all hate me.

Moving on.

There's still the question of how it all works.

What amazes me is how few of you have asked that simple question:

How does the prime counting function work?

Actually one fellow asked, and I refused to answer.

You may have assumed it's because I didn't know myself.

However, if I didn't know how my own prime counting function worked, how do
I keep doing these optimizations, eh?

I've been holding out on you to try and push things.

You see, it seems to me that good 'ol human curiosity will push things in my
favor.

Oh, for those who want a clue, it's about combinations. In fact, the
combination angle may be why the Riemann Hypothesis has a physics
relationship. And that is probably proof that space is not continuous, so
there's likely to be a Nobel prize in all of this, somewhere, for someone
(maybe not, I admit I'm reaching, like usual, sigh).

All I did was figure out how to count combinations of prime numbers by
looking at combinations of linear quadratics, like x+2.

It's really amazingly, simple, breath-takingly beautiful, and incredibly,
incredibly awesome.

I'm trying (really trying!!!) to be a better person just because it's so
incredibly beautiful and awesome.

I feel like I don't deserve this, and it's making me miserable. In fact,
I'm drinking right now.


James Harris


James Harris

unread,
Jul 12, 2002, 8:20:20 PM7/12/02
to

"Dik T. Winter" <Dik.W...@cwi.nl> wrote in message
news:Gz448...@cwi.nl...

> > WilmaScranton wrote:
> In article <3D2DB57F...@yahoo.co.uk> Phil Carmody
<thefatphil_d...@yahoo.co.uk> writes:
> > > From: "James Harris" jst...@msn.com
> > > >With Legendre's (using "floor" instead of my usual "int"),
> > > > 10 - floor(10/2) - floor(10/3) + floor(10/6) + pi(sqrt(10))- 1
gives
> > > > 10 - 5 - 3 + 1 + 2-1 = 4.
> > > >While with my prime counting function,
> > > > 5 - floor((10-4)/6) gives
> > > > 5 - 1 = 4.
> > > >It gets worse, or should I say better?
> ...
> > It becomes likely that James's 'tweak' is a method of pairing terms in
> > the Legendre sum.
>
> No. He just considers odd numbers only.

Really? Do you actually have a clue about how my prime counting function
works, or are you just babbling?

There are *many* key differences between my method and Legendre's.

Not the least of which is Legendre's has the pi(sqrt(N)) term in it. while
mine does not, while it DOES have pi(p_i) terms all through it.

Also, there's a 2^sqrt(N) increase in the number of terms associated with
each prime, with 3 having 1. You see 5 has two, then 7 has 4.

While, in fact it'd be interesting if someone would post it, I'm not sure
how many terms are associated with each prime using Legendre's.

So be useful and post that, if you wish to continue your Quixotic crusade
asserting that my prime counting function is the same as Legendre's.


James Harris


LarryLard

unread,
Jul 12, 2002, 9:58:01 PM7/12/02
to
James Harris wrote:

One for the scrapbook this. I suspect some wall-singing might have occurred
shortly after this post was made.

> Yup, they all hate me.
>
> Moving on.
>
> There's still the question of how it all works.
>
> What amazes me is how few of you have asked that simple question:
>
> How does the prime counting function work?
>
> Actually one fellow asked, and I refused to answer.
>
> You may have assumed it's because I didn't know myself.

I learned in the *playground* not to get riled by people who claimed to know
something, but wouldn't tell.

> However, if I didn't know how my own prime counting function worked,
> how do I keep doing these optimizations, eh?
>
> I've been holding out on you to try and push things.
>
> You see, it seems to me that good 'ol human curiosity will push
> things in my favor.
>
> Oh, for those who want a clue, it's about combinations. In fact, the
> combination angle may be why the Riemann Hypothesis has a physics
> relationship. And that is probably proof that space is not
> continuous, so there's likely to be a Nobel prize in all of this,
> somewhere, for someone (maybe not, I admit I'm reaching, like usual,
> sigh).

s/reaching/bullshitting

> All I did was figure out how to count combinations of prime numbers by
> looking at combinations of linear quadratics, like x+2.
>
> It's really amazingly, simple, breath-takingly beautiful, and
> incredibly, incredibly awesome.
>
> I'm trying (really trying!!!) to be a better person just because it's
> so incredibly beautiful and awesome.
>
> I feel like I don't deserve this, and it's making me miserable. In
> fact, I'm drinking right now.
>
>
> James Harris

Moufang Loop

unread,
Jul 12, 2002, 3:13:48 PM7/12/02
to
James Harris wrote:
>
> ...

>
> The Riemann Hypothesis is just a very informed guess about the nature
> of the distribution of prime numbers i.e. about the prime counting
> function.
>

Do explain to us _what_ the Riemann Hypothesis says and _what_ it
implies about the distribution of prime numbers.

ML

Moufang Loop

unread,
Jul 12, 2002, 3:17:36 PM7/12/02
to
James Harris wrote:
>
>...

>
> Catholics are still dealing with the aftershocks of all the revelations of
> priests who molested children. But I'm sure you all would have certainly
> believe anyone who told you what was going on early on, eh?
>
...


Who gives a damn about child molestation? What traumatizes children is
all the fuss made about a bit of harmless fun.

ML

David C. Ullrich

unread,
Jul 13, 2002, 9:56:50 AM7/13/02
to
On Fri, 12 Jul 2002 11:32:35 -0700, "Dann Corbit" <dco...@connx.com>
wrote:

That may be so. But _most_ of the response to his posts on
this have _not_ consisted of cries that counting primes is
dull. They have instead been replies to his ridiculous
assertions about how the algorithm is the fastest possible
algorithm, how it's totally new, nothing at all like anything
that's been seen before, how it's going to be important in
everyday life for _everyone_, how it leads to a never-stated
differential equation, how it settles the Riemann Hypothesis,
how if Riemann had know about this algorithm he would not
have bothered with his silly hypothesis in the first place,
etc.

If Winter or Kastrup had shared the algorithm, together
with that sort of absurd comment, people would have assumed
they were joking. If they convinced people they were serious
then they would be getting the same derisive jeers that
James has been getting.

>IMO-YMMV.
>--
>C-FAQ: http://www.eskimo.com/~scs/C-faq/top.html
> "The C-FAQ Book" ISBN 0-201-84519-9
>C.A.P. FAQ: ftp://cap.connx.com/pub/Chess%20Analysis%20Project%20FAQ.htm
>


David C. Ullrich

David C. Ullrich

unread,
Jul 13, 2002, 10:02:36 AM7/13/02
to
On Fri, 12 Jul 2002 19:48:28 -0400, "James Harris" <jst...@msn.com>
wrote:

>
[...]


>
>Yup, they all hate me.
>
>Moving on.
>
>There's still the question of how it all works.
>
>What amazes me is how few of you have asked that simple question:
>
>How does the prime counting function work?

When I read this I was about to explain. But you've answered
your own question:

>Actually one fellow asked, and I refused to answer.

In fact more than one person has asked, and you have not
asnwered any of them. In any case you're being funnier than
usual today: People _see_ you ignoring this question, and
then you ask why more people are not asking?

>You may have assumed it's because I didn't know myself.
>
>However, if I didn't know how my own prime counting function worked, how do
>I keep doing these optimizations, eh?
>
>I've been holding out on you to try and push things.
>
>You see, it seems to me that good 'ol human curiosity will push things in my
>favor.
>
>Oh, for those who want a clue, it's about combinations. In fact, the
>combination angle may be why the Riemann Hypothesis has a physics
>relationship.

Note to Corbit: See? This is the sort of thing I was talking about.
If Kastrup or Winter was claiming that the Riemann hypothesis had
something to do with physics, without giving any explanation of
what the connection was, and if they somehow convinced us they
were serious, they'd be getting exactly the same sort of derision
as James has been getting for his ridiculous comments.

>And that is probably proof that space is not continuous, so
>there's likely to be a Nobel prize in all of this, somewhere, for someone
>(maybe not, I admit I'm reaching, like usual, sigh).
>
>All I did was figure out how to count combinations of prime numbers by
>looking at combinations of linear quadratics, like x+2.
>
>It's really amazingly, simple, breath-takingly beautiful, and incredibly,
>incredibly awesome.
>
>I'm trying (really trying!!!) to be a better person just because it's so
>incredibly beautiful and awesome.
>
>I feel like I don't deserve this, and it's making me miserable. In fact,
>I'm drinking right now.
>
>
>James Harris
>


David C. Ullrich

Edward Green

unread,
Jul 13, 2002, 9:38:14 PM7/13/02
to
David C. Ullrich <ull...@math.okstate.edu> wrote in message news:<udc0ju0o6pgffb7p3...@4ax.com>...

> On Fri, 12 Jul 2002 19:48:28 -0400, "James Harris" <jst...@msn.com>
> wrote:
>
> >
> [...]
> >
> >Yup, they all hate me.
> >
> >Moving on.
> >
> >There's still the question of how it all works.
> >
> >What amazes me is how few of you have asked that simple question:
> >
> >How does the prime counting function work?
>
> When I read this I was about to explain. But you've answered
> your own question:
>
> >Actually one fellow asked, and I refused to answer.
>
> In fact more than one person has asked, and you have not
> asnwered any of them. In any case you're being funnier than
> usual today: People _see_ you ignoring this question, and
> then you ask why more people are not asking?

> >You may have assumed it's because I didn't know myself.
> >
> >However, if I didn't know how my own prime counting function worked, how do
> >I keep doing these optimizations, eh?
> >
> >I've been holding out on you to try and push things.

What a pile of shit. Publish your great discovery: if you can.

Paul Sperry

unread,
Jul 14, 2002, 12:37:50 AM7/14/02
to
In article <agn78...@enews2.newsguy.com>,
"Dann Corbit" <dco...@connx.com> wrote:
[...]

> Look everybody, pi(n) calculations are interesting to mathematicians.

No, sorry, but that isn't so. A few mathematicians maybe. Go to your
nearest University library and thumb through a copy of Math. Reviews.
How many of those authors do you think have any interest in counting
primes? Darn few because number theory, let alone counting primes, has
little or nothing to do with their work.

> If not, they ought to be.

Hmm. Sez who?

> If Dik Winter or David Kastrup or someone like that
> had shared the algorithm with us, we would all be keenly enjoying it

No. Not interested [I'm playing it safe, Chris] is not interested. If
Archangel Gabriel fluttered down from on high with a prime counting
algorithm inscribed on a marble tablet the vast majority of
mathematicians would not be interested - at least not in the algorithm.
[...]


> At any rate, these cries of:
> "Calculating pi(x) is dull, dull, dull!"
> are only because it is James.

Who said "dull"? What was said was "not interested".

> IMO

OK

> -YMMV.

It sure does.

RU

unread,
Jul 14, 2002, 2:27:00 AM7/14/02
to
In article <plsperry-0FFBD1...@news-server.sc.rr.com>,
Paul Sperry <plsp...@sc.rr.com> wrote:

>In article <#rKoa1HKCHA.1352@cpimsnntpa03>,
> "James Harris" <jst...@msn.com> wrote:

>> So, why is this function being ignored?
>

> I wonder if you have any idea how
>_little_ interest the mathematics community has in the prime counting
>function.

Yes, but what about the Real World?

What about the importance of prime numbers to the cryptographic
profession?

What about the vast theoretical structure physicists have
painstakingly built that relies on a certain differential equation the
consequences of which are a direct result of the operation (or
something like that) of James's function?

>> The political stakes here are HUGE.
>

>Good one!

But they ARE huge. While no VIPs have as yet commented on James's
breakthrough, I have no doubt that they meet often with their advisors,
discussing what to do when they can no longer pretend things are normal.

Everyone's senses are sharpened. Everyone is holding their breath.

To say nothing of a cynical and basically corrupt professional
mathematical establishment that could --how can I say this politely?--
suffer a drastic reduction in professional credibility when their
cupability in supressing James's innovations becomes widely known.

>> After all, think the next time when you ponder Riemann's Hypothesis,
>

>It's astonishing how little time I spend contemplating the Riemann
>Hypothesis.

Well, you are unusual, then. "Riemann's" has a hold on the popular
imagination. The man who "bags Riemann's" would soon find himself
flying round the globe, pulling down fat speaking fees, surrounded by
beautiful women, invited to parties with the rich and famous.

And people would seek his opinion on whether the taxpayers should
continue to subsidize mathematicians or whether they should just
give up on the current generation of mathematicians and start
afresh.

So if you value your jobs, you will pay attention.

James Harris

unread,
Jul 14, 2002, 11:40:01 AM7/14/02
to

"Paul Sperry" <plsp...@sc.rr.com> wrote in message
news:plsperry-A0AAFA...@news-server.sc.rr.com...

> In article <agn78...@enews2.newsguy.com>,
> "Dann Corbit" <dco...@connx.com> wrote:
<deleted>

> > At any rate, these cries of:
> > "Calculating pi(x) is dull, dull, dull!"
> > are only because it is James.
>
> Who said "dull"? What was said was "not interested".

You got caught Sperry, so quit being defensive. You made a critical post
with the assumption that people would agree with you because, of course, you
are sci.math, right?

However, in that post, you made the statement that mathematicians don't care
about prime counting, and now at least you've gone to saying that YOU don't
care about prime counting.

But, if you don't care, why in the HELL are you still posting?

Why won't you just admit you were wrong, or at least be consistent and QUIT
POSTING in this thread.

I'm sorry but people like you are infuriating in your need to WASTE OTHER
PEOPLE'S TIME.


___JSH


Virgil

unread,
Jul 14, 2002, 4:09:29 PM7/14/02
to
In article <uNnhhy0KCHA.1508@cpimsnntpa03>,
"James Harris" <jst...@msn.com> wrote:

> "Paul Sperry" <plsp...@sc.rr.com> wrote in message
> news:plsperry-A0AAFA...@news-server.sc.rr.com...
> > In article <agn78...@enews2.newsguy.com>,
> > "Dann Corbit" <dco...@connx.com> wrote:
> <deleted>
> > > At any rate, these cries of:
> > > "Calculating pi(x) is dull, dull, dull!"
> > > are only because it is James.
> >
> > Who said "dull"? What was said was "not interested".
>
> You got caught Sperry, so quit being defensive. You made a critical post
> with the assumption that people would agree with you because, of course, you
> are sci.math, right?

He got caught being honest. Programmers are interested in programs.
Mathematicians are interested in mathematics. You have a program.
Guess who will be interested in it and who will to be.


>
> However, in that post, you made the statement that mathematicians don't care
> about prime counting, and now at least you've gone to saying that YOU don't
> care about prime counting.
>
> But, if you don't care, why in the HELL are you still posting?

He is merely pointing out why you are posting to the wrong newsgroup
again.


>
> Why won't you just admit you were wrong, or at least be consistent and QUIT
> POSTING in this thread.

What is he wrong about? Are you talking about your prime counting
program or are you talking about mathematics? If you are talking
about mathematics, you are concealing the subject matter admirably.
If you are talking about a program, talk to programmers in a
programming newsgroup.


>
> I'm sorry but people like you are infuriating in your need to WASTE OTHER
> PEOPLE'S TIME.

When you take so much time to post so many threads irrelevant to
mathematics, your time seems to be cheap enough so that that none of
us need to worry about wasting it.

The ratio of mathematically relevant threads to total threads
started in sci.math by you, JSH, is insignificantly small. Why in
HELL don't you stop wasting the time of those few who still read
your posts, and post onlythat which mathematically relevant.
>
>
> ___JSH
>
>

Paul Sperry

unread,
Jul 15, 2002, 1:17:22 AM7/15/02
to
In article <agr5jk$lp4$1...@idiom.com>, RU <r...@no-spam-idiom.com> wrote:

[Amusing post snipped]

;-)

Paul Sperry

unread,
Jul 15, 2002, 1:21:09 AM7/15/02
to
In article <uNnhhy0KCHA.1508@cpimsnntpa03>,
"James Harris" <jst...@msn.com> wrote:

> "Paul Sperry" <plsp...@sc.rr.com> wrote in message
> news:plsperry-A0AAFA...@news-server.sc.rr.com...
> > In article <agn78...@enews2.newsguy.com>,
> > "Dann Corbit" <dco...@connx.com> wrote:
> <deleted>
> > > At any rate, these cries of:
> > > "Calculating pi(x) is dull, dull, dull!"
> > > are only because it is James.
> >
> > Who said "dull"? What was said was "not interested".

We are not interested in sky diving but we have a hunch it is _not_ dull.

> You got caught Sperry, so quit being defensive. You made a critical post
> with the assumption that people would agree with you because, of course, you
> are sci.math, right?

Huh?
Caught at what?
Defensive about what?
What critical post?
Agree about what?
Of course. Was there any doubt?



> However, in that post, you made the statement that mathematicians don't care
> about prime counting,

We don't think so.

> and now at least you've gone to saying that YOU don't
> care about prime counting.

Conceivably we are a mathematician. (?)



> But, if you don't care, why in the HELL are you still posting?

We are amused.

> Why won't you just admit you were wrong, or at least be consistent and QUIT
> POSTING in this thread.

We are never wrong and I am _never_ inconsistent.

> ___JSH

As a matter of curiosity, why the occasional non "James Harris"
signature?

The Scarlet Manuka

unread,
Jul 15, 2002, 1:02:19 PM7/15/02
to
"James Harris" <jst...@msn.com> wrote:

>All I did was figure out how to count combinations of prime numbers by
>looking at combinations of linear quadratics, like x+2.

Linear quadratics, eh? What will they think of next?

Please do provide us with a definition of "linear quadratics". I'm
sure this concept will revolutionise the whole field of algebra.
Particularly the part where they define the degree of a polynomial.

>In fact, I'm drinking right now.

Really? I never would have guessed.

--
The Scarlet Manuka

James Harris

unread,
Jul 15, 2002, 7:20:32 PM7/15/02
to

"The Scarlet Manuka" <sa...@maths.uwa.edu.au> wrote in message
news:3d330...@news.iprimus.com.au...

> "James Harris" <jst...@msn.com> wrote:
>
> >All I did was figure out how to count combinations of prime numbers by
> >looking at combinations of linear quadratics, like x+2.

OOPS!


> Linear quadratics, eh? What will they think of next?
>
> Please do provide us with a definition of "linear quadratics". I'm
> sure this concept will revolutionise the whole field of algebra.
> Particularly the part where they define the degree of a polynomial.

Ok, so I was thinking one thing and typed another. At least I posted x+2.
I doubt that anyone but my diehard (and strangely obsessive, and vicious)
critics would debate whether or not I know what a quadratic is.

Then again, the sci.math newsgroup is VERY weird on this sort of thing, as
they like to STILL mention similar goof statements of mine from YEARS back.

LOL. I've found it humorous for quite some time now.


> >In fact, I'm drinking right now.
>
> Really? I never would have guessed.

Um, unfortunately, I was drinking a lot. In fact, some of my, um, most
vituperative posts, ok, most of them have been written under the influence.

It seems that a couple of beers gets me very upset about the mathematical
establishment.


James Harris


David Kastrup

unread,
Jul 15, 2002, 8:49:31 PM7/15/02
to
"James Harris" <jst...@msn.com> writes:

> Um, unfortunately, I was drinking a lot. In fact, some of my, um,
> most vituperative posts, ok, most of them have been written under
> the influence.
>
> It seems that a couple of beers gets me very upset about the
> mathematical establishment.

It might help a lot to your reception here if you were restricting
voicing your opinions on "the establishment" when you can be certain
that no beer has been involved in their creation. Of course, this
might improve the quality for other types of post, too.

I would estimate that most posts with sentences starting with "Folks,
..." would fall under that category.

Randy Poe

unread,
Jul 16, 2002, 9:02:29 AM7/16/02
to
James Harris wrote:
> Ok, so I was thinking one thing and typed another. At least I posted x+2.
> I doubt that anyone but my diehard (and strangely obsessive, and vicious)
> critics would debate whether or not I know what a quadratic is.

Probably you do. But at times you want to include infinite power
series under "polynomials", as well as things with fractional
powers. So I'm not sure how extensive you knowledge of polynomials
is.

> Then again, the sci.math newsgroup is VERY weird on this sort of thing, as
> they like to STILL mention similar goof statements of mine from YEARS back.

You don't have to go back years for the above.

> It seems that a couple of beers gets me very upset about the mathematical
> establishment.

I don't recall a single post of yours that did not express
this sentiment. Does that mean that all of your posts have
been beer-inspired?

- Randy

0 new messages