[seqfan] Re: Dead b-files for A127816 and others

Hagen von Eitzen math at von-eitzen.de
Thu Jul 23 00:09:58 CEST 2009

Richard Mathar schrieb:
> I am putting short-term replacements a127817.txt and a127816.txt under
> http://www.strw.leidenuniv.nl/~mathar/progs/a127816.txt
> http://www.strw.leidenuniv.nl/~mathar/progs/a127817.txt
> with a comment at the top to indicate the quality of the files--
> which is the count of -1 entries in the 100,000 entries.
> The files are still being updated. Experience shows that my
> speed of compatations is not very competitive, so everybody is
> welcome to submit files of higher quality than these.
It seems that extending the brute-force limit further does not make much 
sense once the number of gaps has become small enough.
Instead I almost routinely tackle the gaps specially, with threaded 
brute force nicely sped up by a bit of number theory.
For example,
A128372(49) = 46528545441593 (that was a toughy!),
A078457(34) = 622699582951,
A078457(74) = 337148405497007,
A078457(160) = 66190429538797
would hardly have been found by brute force.

This way I could report e.g.
yesterday. Just checking my mail, it seems but that my yesterday 
submission for
A127817(8) = 2592842671511
has got lost or I managed to miss the submit button - I'll check again 
(Note that both these findings confirm a conjectured value found by Joe 
Crump long ago).

Also I had quickly found by my method
A127816(49) = 161903501
A127817(50) = 393217991
A127817(56) = 276250159,
which may already have been in the "lost" a-files.


More information about the SeqFan mailing list