Is this interesting enough to submit?

[N. J. A. Sloane]

I don't disagree with Dr. Neil J. A. Sloane.  I am not submitting the
sequence of which I asked.

But A002496 differs from almost all of the "couple of thousand
sequence lik[e]  that begin Primes of the form ******"

in that:

(1) It is perhaps the most important such sequence for which we do not
know if it is finite or infinite;
(2) see, for instance,   R. K. Guy, "Unsolved Problems in Number
Theory", 3rd edition, A2;
(3) njas judged the seq to be "nice";
(4) it is one of the four "unattackable" problems mentioned by Landau
in the 1912 Fifth Congress of Mathematicians in Cambridge. The four
were
1. The Goldbach conjecture,
2. Twin prime conjecture,
3. Legendre's conjecture that for every n there exists a prime p
between n^2 and (n+1)^2 (Hardy and Wright 1979, p. 415; Ribenboim
1996, pp. 397-398), and
4. The conjecture that there are infinitely many primes p of the form
p=n^2+1 (Hardy and Wright 1979, p. 19; Ribenboim 1996, pp. 206-208).
Weisstein, Eric W. "Landau's Problems." From MathWorld--A Wolfram Web
Resource. http://mathworld.wolfram.com/LandausProblems.html

I don't disagree that it might be a bad precedent to accept partial
sums of this, as it would be pretty much downhill to do the same for
the other couple of thousand.

I saw no harm in swimming at the shallow end of one of Landau's four
"unattackable" problems.  The partial sum transform is so elementary
that it was unlikely to shed any light on the problem.

I wonder if there are more seqs clustered around the "nice" and "core"
seqs than around random seqs.  I hope so.

I adhere to following the taste of njas, and see no harm in asking
whether a seq is interesting.  If "the usual suspects" did this, there
would be fewer "less" and "uned" seqs in the OEIS.

Best,

Jonathan Vos Post


On 6/9/08, N. J. A. Sloane <njas at research.att.com> wrote:
> -----Original Message-----
>  From: Jonathan Post <jvospost3 at gmail.com>
>
>  ...
>
>  Is this interesting enough to submit?
>
>
> Partial sum of A002496 Primes of form n^2 + 1.
>
>  ...
>
>  Me:  No!
>
>  (There are a couple of thousand sequence lik
>
>  that begin Primes of the form ******
>
>  and there's no reason to include all their
>
>  partial sums)
>
>  Neil
>




bmk> From seqfan-owner at ext.jussieu.fr  Thu Jun  5 11:20:52 2008
bmk> Date: Thu, 5 Jun 2008 19:19:31 +1000
bmk> From: Brendan McKay <bdm at cs.anu.edu.au>
bmk> To: Joerg Arndt <arndt at jjj.de>
bmk> Cc: Jonathan Sondow <jsondow at alumni.princeton.edu>, seqfan at ext.jussieu.fr
bmk> Subject: Re: 2 things about the OEIS Welcome Page
bmk> 
bmk> The reason that "subscribe" and similar features do not work is that
bmk> the ATT server does not include the Last-Modified: header in serving
bmk> OEIS pages. This is not the case for all pages on that server,
bmk> so maybe there is some way to do it. The best chance would be an
bmk> ordinary URL (without the "?q..." stuff) for the raw text version of the sequence. Is there one?
bmk> 
bmk> Btw, since wget is mentioned I'll mention that "wget -S URL" is one
bmk> way to see the HTTP headers.
bmk> 
bmk> Brendan.

The HTML lines  of the format
would have to be inserted by the program (cgi, perl, php, ....) which creates
the individual HTML page. The web server does not interfere with that.

This functionality needs some effort since the information on the date
(and time) of the last change is not an explicit datum in the database;
lines for strings of the form "(Jan|Feb|Mar|....) (0-2)(0-9) 200(0-9)", 
occasionally wrapped around into a new line, sort these on a per-A-number
basis according to the dates, and then copy the last output of this filter
into the "last-modified" line of the HTML header.
explicit day tags aside each new comment, formula, extensions etc. In a
UNIX-style approach, these cases could default to 1 Jan 1970 or perhaps to
Neil's birth date -:).

Richard



Dear Richard, Joerg, Brendan and other seqfans,

Thanks for all your suggestions on "subscribing" to an OEIS entry.
Unfortunately, I don't understand technical things like wgets, HTML, etc. So
I still don't know how to subscribe without getting false alarms. :(

Jonathan Sondow 


On 6/10/08 10:40 AM Richard Mathar wrote:

> 
> bmk> From seqfan-owner at ext.jussieu.fr  Thu Jun  5 11:20:52 2008
> bmk> Date: Thu, 5 Jun 2008 19:19:31 +1000
> bmk> From: Brendan McKay <bdm at cs.anu.edu.au>
> bmk> To: Joerg Arndt <arndt at jjj.de>
> bmk> Cc: Jonathan Sondow <jsondow at alumni.princeton.edu>, seqfan at ext.jussieu.fr
> bmk> Subject: Re: 2 things about the OEIS Welcome Page
> bmk> 
> bmk> The reason that "subscribe" and similar features do not work is that
> bmk> the ATT server does not include the Last-Modified: header in serving
> bmk> OEIS pages. This is not the case for all pages on that server,
> bmk> so maybe there is some way to do it. The best chance would be an
> bmk> ordinary URL (without the "?q..." stuff) for the raw text version of the
> sequence. Is there one?
> bmk> 
> bmk> Btw, since wget is mentioned I'll mention that "wget -S URL" is one
> bmk> way to see the HTTP headers.
> bmk> 
> bmk> Brendan.
> 
> The HTML lines  of the format
> <meta http-equiv="Last-Modified" content="Thu, 8 Nov 2007 00:00:00 GMT">
> would have to be inserted by the program (cgi, perl, php, ....) which creates
> the individual HTML page. The web server does not interfere with that.
> 
> This functionality needs some effort since the information on the date
> (and time) of the last change is not an explicit datum in the database;
> so the script which creates the HTML page would have to scan the database
> lines for strings of the form "(Jan|Feb|Mar|....) (0-2)(0-9) 200(0-9)",
> occasionally wrapped around into a new line, sort these on a per-A-number
> basis according to the dates, and then copy the last output of this filter
> into the "last-modified" line of the HTML header.
> This may not work with old OEIS entries because these are not carrying the
> explicit day tags aside each new comment, formula, extensions etc. In a
> UNIX-style approach, these cases could default to 1 Jan 1970 or perhaps to
> Neil's birth date -:).
> 
> Richard
>