uned sequences A071339 - A071341

[Richard Guy]

I agree with the recent reMarcs.    R.

On Mon, 9 Jun 2003, Marc LeBrun wrote:

>  >=David Wasserman
>  > I would suggest deleting these three, but first I'll ask if anyone 
> disagrees.
> 
> Please allow me to very respectfully disagree, in the strongest possible terms.
> 
> This issue arises periodically, in part as a natural consequence of the 
> workload the OEIS entails.  However it strongly bears not only on the 
> editorial burdens, but vitally on the ultimate long-term value of the OEIS.
> 
> As editors we should strive to improve accuracy, completeness, coherence, 
> etc, but not drift into acting as gatekeepers, except where the utility of 
> the OEIS would clearly be adversely impacted by a submission.
> 
> Bear in mind that the OEIS is *not* a journal, it's a reference database, 
> and, moreover, one that is used by programs, such as superseeker, as well 
> as humans.
> 
> As time passes, the frequency, depth and importance of these automated 
> searches can only increase.  This has direct bearing on the "appreciation" 
> of the investment people make by contributing to the OEIS.
> 
> Having lots of sequences can enable programs to mine very valuable results 
> that we could not envision at the time of submission.
> 
> Conversely, "censored" sequences are often not likely to be effectively 
> resynthesized by such tools, simply because there are so many viable forms 
> such derivations might take.
> 
> The potential value of including a possibly marginal sequence greatly 
> outweighs the slight incremental cost.  "A numbers" are not a scarce resource.
> 
> Therefore sequences should be rejected only for the very strongest reasons 
> (such as errors, incomprehensible descriptions, etc).
> 
> It does not, however, seem sufficient grounds to reject a sequence simply 
> because someone, with enough cleverness and insight, might able to derive 
> it from another sequence.  This *is* a reasonable criteria for, say, a 
> submitted journal article with a "trivial" or "known" result, but not the OEIS.
> 
> Indeed, there are a number of existing sequences that differ only by a 
> single, arguable, boundary term, or by shifting offset or other simple 
> variations.  Yet these details can make a critical difference in whether or 
> not an automatic transform will be able to "hit" on interesting results in 
> the future (eg consider how shifting the offset by one radically affects a 
> Mobius transform, say).
> 
> The advertised goal of the OEIS is to include "ALL interesting 
> sequences".  Evidently *someone* was interested enough to go to the trouble 
> of submitting the sequence, manifestly so if they had to, say, write a 
> program, or undertake some analysis, in order to generate the submitted 
> information.  It is important to respect the submitters' level of interest 
> and motivation, even when we aren't as enthusiastic ourselves.
> 
> Certainly there are "spam" sequences, or lunatic submissions, or cases 
> where the submitter obviously got carried away using power tools to 
> generate excessively many too-similar variants, or the like.  But these are 
> pretty blatant.  If there's much room for reasonable people to disagree, a 
> sequence really should be presumed acceptable, unless proven otherwise.
> 
> In this particular case, while I'd be happy to see the FORTRAN code 
> sequestered off elsewhere under some link, all the submitted sequences look 
> reasonable, and I believe their inclusion would enrich the "core" sequence 
> A004018, as well as the OEIS as a whole.
> 
> Alas this doesn't ease the editorial burden while furthering this wonderful 
> enterprise.  Your efforts are greatly appreciated.
> 
> Thanks!
> 
>